Skip to main content

Full text of "Lowie's selected papers in anthropology"

See other formats


Digitized by the Internet Archive 

in 2011 with funding from 

LYRASIS Members and Sloan Foundation 



LOWIE'S Sekded 
Papers in Anthropology 


Berkeley and Los Angeles • 1960 

University of California Press 

Berkeley and Los Angeles, California 

Cambridge University Press, London, England 

© 1960 by The Regents of the University of California 

Library of Congress Catalog Card Number: 60-8314 

Printed in the United States of America 

Designed by Marion Jackson Skinner 




Editor ^s Preface 

The selection of essays in this volume has been 
largely determined by Robert Lowie himself. He not only com- 
piled the syllabus of a seminar on his own works (Appendix) but 
also, during the last ten years of his life, returned again and again 
to an appraisal of his work and the intellectual climate in which it 
was pursued. The "Autobiographical Data" in this volume, chap- 
ters in Robert H. Lowie, Ethnologist: A Personal Record (Uni- 
versity of California Press, 1959), and a number of brief notes 
that he sent to some of his former students testify to the value he 
set on his various writings. In 1948 I received the following note 
from him: 

An author is notoriously disqualified from properly assessing 
what he has written. Nevertheless, it is of some psychological interest 
for others to learn what he himself values as most significant in his 
output. The award of the Viking Fund medal in December, 1947, 
prompted me to survey my writings with an eye to determining what 
I could myself regard as "contributions," i.e., as reasonably "original" 
and formally acceptable additions to anthropological thinking, as 
things not said at all, or not so well, by others. The results are 
meager, but possibly suggestive. 

The Plains Indian Age Societies (AMNH-P 11:955-984, 1916) 

* Family and Sib (AA 21:28-40, 1919) 

* A Note on Aesthetics ( AA 23: 170-174, 1921 ) 
Primitive Religion, pp. 3-32, 1924 

* A Note on Relationship Terminologies (AA 30:263-267, 1928) 
The Crow Indians, pp. 104-118, J 935 

* Studies in Plains Indian Folklore (UC 40:1-28, 1942) 

* A Case of Bihngualism (Word, 1:249-259, Dec. 1945) 

An asterisk indicates that the paper is included in this volume. 

Editors Preface 

However clear Lowie made the appraisal of his own writing, 
an editor is nevertheless constrained to exercise selective judg- 
ment, particularly when the author's works are as voluminous and 
diversified as those of Lowie. In this task A. L. Kroeber, Paul 
Radin, and Leslie Spier gave generously of their advice, and I 
have included all the articles on which there was consensus. 
Practical considerations nevertheless remained. While trying to 
choose articles for their excellence and as illustrations of the 
range and development of Lowie's thoughts on anthropology, I 
was limited by considerations of length and availability. Thus, im- 
portant representative papers that are easily found elsewhere are 
not reprinted here. Also, the arbitrary decision was made not to 
reproduce portions of his books and longer monographs. Ob- 
viously, no one can understand fully the directions of Lowie's 
thinking without reference to writings not here included, and 
above all, to his ten major books and to the eighteen major 
monographs on the Plains Indians published primarily by the 
American Museum of Natural History. For example, the omis- 
sion here of Lowie's contributions to primitive religion represents 
a serious lacuna that the reader can fill by turning, inter alia, to 
his Frimitive Religion (1924). This collection therefore cannot 
pretend to be an intellectual history. I hope, however, that it 
is a fair sample of Lowie's contributions to anthropology. 

In the same vein, another comment on the contents of this 
volume is warranted. As Kroeber pointed out in his obituary of 
Robert Lowie written for Sociologus (n.s.. Vol. VIII, No. 1 [1958], 
1-3 ) , "Perhaps a tenth of Lowie's 300 titles — apart from probably 
some additional 200 reviews — lie outside anthropology in sub- 
ject — on philosophy, general science, belles-lettres, aesthetics, or 
biography. A few samples of his interests are Feuerbach, Spencer, 
Tolstoy, Haeckel, feminism, authority in science, mysticism, 
Wundt, the golden section, skepticism, bilingualism ( he was him- 
self one of the rare complete bilinguals), letters from Mach, pa- 
rochialism," Although I would not draw the boundaries of anthro- 
pology as closely as Kroeber does in this statement, nevertheless 
it is true that Lowie's wide-ranging interests, and particularly 
his early interest in philosophy are barely represented even tliough 
they are of major importance to an understanding of his scien- 


Editors Preface 

tific convictions. There was an embarrassment of riches. Since 
Lowie himself did not include his earlier writings on philosophy 
among his contributions to anthropology, they are omitted here. 
Nevertheless, the only two professional photographs that hung 
over the fireplace in his study at the time of his death were of 
Mach and Haeckel. 

Lastly, in respect to the contents of this volume, the presence 
of six hitherto unpublished papers should be noted. These were 
found in Lowie's files after his death in September, 1957. They 
were made available by Luella Cole Lowie. Although authors 
are not always well served by posthumous publications of manu- 
script materials, it was deemed desirable to make these essays 
available in this volume. 

Over and above the question of selection, there remains the 
question of editing. Lowie does not have the privilege of wield- 
ing his own blue pencil as, for example, Kroeber did in The Nature 
of Culture (Chicago: University of Chicago Press, 1952) or as 
Hallowell did in Culture and Experience (Philadelphia: Univer- 
sity of Pennsylvania Press, 1955 ) . To presume to perform this func- 
tion for Lowie would be an impertinence. Therefore the papers 
brought together here are in the form in which they were first 
written. Undoubtedly, Lowie would have deleted, revised, and 
elucidated much that is here reproduced. Had he done so, this 
volume would have been much improved. 

No attempt has been made to standardize the spelling of tribal 
names, regions, or other words which in Lowie's manuscripts 
varied over the years. Nor has there been any tampering with 
certain stylistic idiosyncrasies. Appended lists of references have 
been omitted in the interest of brevity. The original footnotes in 
each article have, however, been retained, and some attempt has 
been made to standardize them for this volume as a whole. 

Acknowledgments are always a pleasure. The debt to Kioeber, 
Radin, and Spier has already been mentioned, although patently 
final responsibility must rest with the editor. Walter Goldsclmiidt 
initiated the idea of this collection of essays, persuaded me to 
undertake their selection, and made arrangements for their pub- 
lication. Luella Cole Lowie has been judicious, devoted, and 
indefatigable in her labors as Lowie's literary executor. Without 


Editor s Preface 

her generous collaboration, my task would have been heavy. Miss 
Cynthia Nelson of the Department of Anthropology of the Uni- 
versity of California, Berkeley, supplied some necessary biblio- 
graphical data for the footnotes and Miss Dorothy H. Huggins of 
the University of California Press undertook the Herculean task 
of standardizing them throughout the volume. The scholarly 
benevolence of the editors of various journals who allowed articles 
to be republished is gratefully acknowledged, and the following 
publishers and copyright holders are especially thanked for per- 
mission to use the material reprinted here: American Anthropolo- 
gist, American Folklore Society, Columbia University Press, The 
Michigan Academy of Science, Arts, and Letters, National Acad- 
emy of Sciences, Royal Antliropological Institute, University of 
Chicago Press (publisher of American Journal of Sociology), Uni- 
versity of Michigan, Yale Law Journal. 

To all these Robert Lowie also would have rendered generous 
thanks. The editor alone must bear the blame for any shortcom- 

V CoRA^ Du Bois 

February, 1959 
Center for Advanced Study 
in the Behavioral Sciences 



Autobiographical Data by 
Robert H. Lowie 

PART I Kinship and Social Organization 

^ 1 Social Organization ( 1914) 17 

2 Exogamy and the Classificatory Systems 
of Relationship (1915) 48 

3 Historical and Sociological Interpreta- 
tions of Kinship Terminologies (1916) 65 

4 The Kinship Systems of the Crow and 
Hidatsa (1917) 75 

5 Family and Sib ( 1919 ) 82 

, 6 A Note on Relationship Terminologies y- 

(1928) 95 

7 The Omaha and Crow Kinship Termi- 
nologies (1934) 100 

8 The Family as a Social Unit ( 1933 ) 111 

9 Nomenclature and Social Structure 128 





PART II Literature, Language, and 

yiJO A Note on Aesthetics ( 1921 ) 137 

11 Some Cases of Repeated Reproduction 

/ (1942) 143 

y^ ^ 12 A Case of Bilingualism (1945) 154 

/ 13 Observations on the Literary Style of the 

Crow Indians (1950) 165 

14 Evolution and Diffusion in the Field of 

Language 177 

PART III Relation of Ethnology to Other 

V'^ 15 Psychology and Sociology ( 1915) 189 

1 6 Oral Tradition and History ( 1917 ) 202 

J J J -^— J 7 Psychology, Anthropology, and Race 

(1923) 211 

18 Incorporeal Property in Primitive Society 

(1928) 225 

'^y^\19 Economic Factors and Culture 240 

20 Property Rights and Coercive Powers of 
Plains Indian Military Societies (1943) 247 

21 Some Aspects of Political Organization 
Among the American Aborigines ( 1948 ) 262 

PART IV Theories and Theorists 

22 A New Conception of Totemism ( 1911 ) 293 


23 On the Principle of Convergence in Eth- 
nology (1912) 312 

24 Ceremonialism in North America ( 1914) 336 

^ 25 Edward B. Tylor (1917) 365 

^ xj 26 Lewis H. Morgan in Historical Perspec- 

. tive (1936) 372 

27 Cultural Anthropology: A Science 

(1936) 391 

\/ 28 Evolution in Cultural Anthropology: A 

Reply to Leslie White ( 1946 ) 411 

\ 29 Franz Boas, 1858-1942 (1947) 425 

30 Some Problems of Geographical Distri- 
bution (1951) 441 

31 Contemporary Trends in American Cul- 
tural Anthropology ( 1955 ) 461 

' 32 Berthold Laufer as Ethnologist 472 

V, 33 The Development of Ethnography as a 

Science 480 


Syllabus of Seminar on Work of Robert 

H. Lowie 495 




PART II Literature, Language, and 

10 A Note on Aesthetics ( 1921 ) 137 

11 Some Cases of Repeated Reproduction 

/ (1942) 143 

yC -' 12 ACaseof Bilingualism (1945) 154 

13 Observations on the Literary Style of the 
Crow Indians (1950) 165 

14 Evolution and Diffusion in the Field of 
Language 177 

PART III Relation of Ethnology to Other 

^~ 15 Psychology and Sociology ( 1915) 189 

1 6 Oral Tradition and History ( 1917 ) 202 

J J J -^^ J 7 Psychology, Anthropology, and Race 

(1923) 211 

18 Incorporeal Property in Primitive Society 

(1928) 225 

'w/^\i9 Economic Factors and Culture 240 

20 Property Rights and Coercive Powers of 
Plains Indian Military Societies (1943) 247 

21 Some Aspects of Political Organization 
Among the American Aborigines ( 1948 ) 262 

PART IV Theories and Theorists 

22 A New Conception of Totemism ( 1911 ) 293 


23 On the Principle of Convergence in Eth- 
nology (1912) 312 

24 Ceremonialism in North America ( 1914) 336 

^ 25 Edward B. Tylor (1917) 365 

X ^ 26 Lewis H. Morgan in Historical Perspec- 

/ tive (1936) 372 

27 Cultm'al Anthropology: A Science 

(1936) 391 

\/ 2H Evolution in Cultm'al Anthropology: A 

Reply to Leslie White ( 1946 ) 411 

\ 29 Franz Boas, 1858-1942 (1947) 425 

30 Some Problems of Geographical Distri- 
bution (1951) 441 

31 Contemporary Trends in American Cul- 
tural Anthropology ( 1955 ) 461 

32 Berthold Laufer as Ethnologist 472 

^ , 33 The Development of Ethnography as a 

Science 480 


Syllabus of Seminar on Work of Robert 

H. Lowie 495 




Abbreviations in Footnotes 

AAA American Anthropological Association 

AMNH American Museum of Natural History 

BAE Bureau of American Ethnology 

PMH Peabody Museum of American Archaeology and Eth- 
nology, Harvard University 

RAI Royal Anthropological Institute of Great Britain and 


USNM United States National Museum 


Autobiographical Data by 
Robert H. Lowie 

The circular letter of July 22, 1943,* calls for a state- 
ment about "Discoveries which you regard as the most important, 
and the circumstances under which they were made." Contribu- 
tions in different fields of knowledge are probably not commen- 
surable; accordingly I feel the need of interpreting the request 
rather broadly, — as a demand for the writer's subjective evalua- 
tion of his more significant scientific activities. For the sake of 
convenience I shall divide the account under the heads of Field 
Work; Primitive Social Institutions; Primitive Art and Literature; 
Primitive Religion; South American Ethnography; Theoretical 
Position. I shall conclude with some General Remarks. 

Field work. — Given primitive cultures that have not yet been 
adequately described, any trained observer sent to study them is 
able to make "discoveries." In this respect I was extremely lucky. 
The first tribe I visited, the Lemhi Shoshone of Idaho ( 1906), and 
their cognates in Utah and Nevada, to whom I paid some atten- 
tion in 1912 and subsequently, were so little known that any 

* From the National Academy of Sciences. This paper is Lewie's reply to the 
circular letter. 

Autobiographical Data 

tidbit of information was welcome. In consequence the sparse 
data I presented in The Northern Shoshone (AMNH, Anthro- 
pological Papers, II [1909], 165-306) and Notes on Shoshonean 
Ethnography (ibid., XX [1924], 187-314) assumed a significance 
quite disproportionate to their intrinsic worth, for they did assist 
in more clearly visualizing the relations of the Basin to the 
Califomian and to the Plains areas, respectively. My maiden trip 
likewise suggested my doctoral dissertation. The casual way in 
which Lemhi narrators interwove explanatory elements into their 
tales convinced me that they were not an organic part of the 
stories. This idea fell in with one of Boas's pet principles, and he 
encouraged me to elaborate it, whence my first theoretical paper, 
"The Test Theme in North American Mythology" (Journal of 
American Folk-Lore, XXI [1908], 97-148). 

I was also singularly lucky in my studies of the Crow Indians, 
begun in 1907, continued regularly in the summers from 1910 to 
1916, and once more in 1931. Here there had been no information 
above amateur calibre excepting Lewis H. Morgan's data on the 
social organization and kinship system. His report of matrilineal 
clans had been discountenanced in the wave of general skepticism 
that dominated the turn of the century, but I was soon able to 
vindicate its accuracy. Further, thanks to the amiability of the In- 
dians as well as to an extraordinarily competent and faithful in- 
terpreter, I came to accumulate a good deal of material on social 
institutions, ceremonies, and religious beliefs. The results ap- 
peared in various papers listed in a compact book that sum- 
marized them, — The Crow Indians (New York, 1935). 

Incidental comments, in writing or orally, by Bronislaw Mali- 
nowski, Leslie Spier, W. I. Thomas, and others suggest that the 
data were appreciated from various points of view, perhaps par- 
ticularly as helping to elucidate the influence of an ideological 
pattern on individual human behavior. However, I cannot honestly 
accept Wissler's flattering reference to my work as an example of 
"approximately complete data" on an aboriginal culture (Clark 
Wissler, Man and Culture, New York, 1923, p. 51), for I am 
painfully aware of my neglect of the economic and technological 
aspects of life. 

Two phases of my Crow work may perhaps be dignified as 
"discoveries." Although the literature on American Indians teemed 

Autobiographical Data 

with references to visionary experiences, there were never more 
than a handful from any one tribe, — apart from those recounted 
in myth. Thanks to the friendhness of my informants, I was able 
to get a relatively large number of such accounts from Crow 
visionaries. These, under the stimulus of a passage in Harald 
Hoffding's The Philosojyhy of Religion, I subsequently analyzed 
in my Primitive Religion (New York, 1924) from the point of 
view of individual differences. As a graduate student I had been 
fascinated by Galton's Inquiries into Human Faculty, and Boas 
had in his lectures emphasized the variability of human beings in 
primitive societies. So far as I can see, I went beyond my prede- 
cessors in suggesting that the psychological experiences in ques- 
tion probably reflected auditory, kinaesthetic, visual, motor, or 
other peculiarities of my informants. I do not know to what ex- 
tent the point attracted notice. If I remember correctly, it ap- 
pealed to R. R. Marett,who wrote me a commendatory note about 
the book; and it has recently been cited by Spier ("Franz Boas 
and Some of his Views," Acta Americana, I [1943], 119). 

The second "discovery" relates to military organizations of the 
same people. Earlier reports of corresponding institutions, mainly 
among neighboring tribes of the Plains, had been worked into 
Heinrich Schurtz's comprehensive scheme of associations, in 
which the Crow (and other) military societies represented an 
evolutionary stage between "age-societies" and "clubs." Brief 
work among the Assiniboine Indians in 1908 had made me doubt 
the age qualification for membership as more than incidental, 
and among the Crow I systematically pumped every male I could 
interview concerning his former participation in the organizations 
under discussion. It thus turned out that within the remembered 
period the associations investigated were unquestionably not 
constituted by age-mates. An extension of this line of thought 
will be noted below. 

As a natural corollary to my Crow work, Wissler sent me to 
the related Hidatsa. Though I never published my notes on them 
in full, the comparison of these two Plains Indian groups, ob- 
viously sprung from a single ancestral tribe some centuries ago, 
led to various surmises concerning the course of development 
each had taken since the separation. 

Other field work will be referred to under the topical headings. 

Autobiographical Data 

Primitive social institutions. — My discussion of kinship ter- 
minologies was important in the sense that it ushered in a re- 
vival of interest in the topic, from which most American students 
had been alienated by Morgan's evolutionary scheme. The sub- 
ject had always bored me, but W. H. R. Rivers' Kinship and Social 
Organisation (1914) convinced me that here was a field in which 
rigorous formulation of problems was possible. Among other 
things, Rivers had suggested a correlation between a clan system 
and one type of "classificatory" kinship nomenclature, an idea an- 
ticipated by Tylor with reference to moieties. Rivers' conception 
rested on Oceanian data; I examined available North American 
information, confirming Rivers' general point ("Exogamy and the 
Classificatory System of Relationship," American Anthropologist, 
XVII [1915], 223-239.^ Finding enormous gaps in the relevant 
data, however, I turned propagandist, advertising the significance 
of the problem on every occasion. The result was the accumula- 
tion of much trustworthy material by Kroeber, Sapir, Speck, Elsie 
Clews Parsons, Gifford, Spier, Lesser, and others. Specifically, the 
Southwestern terminologies, hardly known before, came to be 
described in great abundance. 

My own role in this development, apart from that of pro- 
moter, was a restricted one. I grew more open-minded con- 
cerning the part played by the levirate and sororate ( championed 
by Sapir ) and tried to link the origin of the North American clans 
with particular groupings of kin ("Family and Sib," American 
Anthropologist, XXI [1919], 28-40 ).2 Under the category of "dis- 
covery" I should place my findings on the system of the Hopi 
Indians. Since this tribe was the only Shoshonean people with a 
clearly developed clan organization, they provided an ideal field 
for testing Rivers' theory. It appeared that they really differed 
radically from their clanless congeners and precisely so as to sug- 
gest a correlation with the clan structure (Hopi Kinship, AMNH, 
Anthropological Papers, XXX [1929], 361-388). 

My personal contributions, however, remained of limited scope. 
What attracted me were the functional relations and to some ex- 
tent historical connections suggested by kinship terminologies. I 

^ Reprinted as No. 2 in this volume. ^ 

^ Reprinted as No. 5 in this volume. 

Autobiographical Data 

never did what might reasonably have been expected of me, viz., 
bring Morgan up to date even for a single major area, as Gifford 
did for California. Radcliffe-Brown, Kroeber, Gifford, Lesser, 
Warner, and Aginsky have all probably been more deeply in- 
terested in the relevant facts than I ever was. It strikes me as 
characteristic that I turned to the conceptual aspect of the field, 
proposing a new typology of kinship systems to supersede Mor- 
gan's and Rivers' ( "A Note on Relationship Terminologies," Amer- 
ican Anthropologist, XXX [1928], 263-267).^ Its essential sound- 
ness seems indicated by the fact that soon after its publication 
Kirchhoff independently elaborated a virtually identical scheme. 
It also stimulated some more elaborate attempts by Warner and 

By and large, my work on the Plains Indian military organiza- 
tions probably constitutes my most important "discovery." As 
stated, it grew out of my early field work and a critical reading 
of relevant theories. Coupled with my doubts concerning the 
age factors was the conclusion that here, too, Boas's principle of 
secondary association could be abundantly demonstrated. These 
ideas were adumbrated in my descriptive report on The Assini- 
hoine (AMNH, Anthropological Papers [1909], 75-98). They 
aroused favorable comment from Sapir and others, and Clark 
Wissler was stimulated to organize one of the Museum's major 
projects, a systematic study of Plains societies in which virtually 
all members of the staff participated in some measure for half 
a dozen years. The results were embodied in Volume XI of the 
Museum series, the concluding section "Historical and Compara- 
tive Summary" (New York, 1916) being naturally entrusted to 
me. This paper met with more favor than any other of my techni- 
cal writings. Boas, generally chary of praise, told me that it 
seemed to be "exceedingly well done," and it is the only one of 
my publications, I think, of which Goldenweiser wholeheartedly 

In retrospect I should discriminate. There is, for my present 
taste, too much of what Radcliffe-Brown calls "conjectural his- 
tory," and to many of the specific chronological hypotheses I now 
attach little importance. Nevertheless some of the historical 

* Reprinted as No. 6 in this volume. 

Autobiographical Data 

results still appear well grounded. I established a closer relation 
at one time between the Arapaho and the Village tribes than ap- 
pears from documentary evidence; and I demonstrated the iden- 
tity of the Arapaho Tomahawk society with the Lumpwoods of 
the Hidatsa. However, the principal conclusions seem to me to lie 
in other directions. I think I proved that during the historical 
period the very organizations which most clearly seemed to ex- 
emplify Schurtz's age-sets did not at bottom correspond to particu- 
lar age-levels, membership primarily hinging on payment of a fee. 
They were age-classes only because age-mates tended to buy 
membersliip jointly; without the payment entrance was impos- 
sible, and on the other hand a man who had purchased his mem- 
bership retained it irrespective of age until he had sold what were 
basically property rights. Thus, the age factor turned out to play 
a very different part from that in Schurtz's scheme. Secondly, in 
contravention of unilinear schemes of evolution, I showed that 
the superficially similar age-sets of the Masai and clubs of Mela- 
nesia were radically different from the Plains Indian associations, 
each area having produced a phenomenon sui generis. 

Before leaving the subject of Social Organization I feel im- 
pelled to comment on my Primitive Society (New York, 1920), 
whose success illustrates that an opportune publication may 
create an effect quite disproportionate to its merit. There were 
indeed a few voices that damned with faint praise. Kroeber 
(American Anthropologist, XXII [1920], 377-381) commended 
its honesty, but emphasized the lack of "broad results," the "com- 
parative sterility," want of "visions of more ultimate enterprise." 
Rivers (ibid., 278-283) criticised my overcautious and "mechani- 
cal" use of diffusion and rightly pointed out my ignorance of 
Oceanian and African data. But the discordant voices were 
drowned in a chorus of acclamation. Boas, meeting me soon after 
the book had appeared, said, "I think you have written an awfully 
good book"; Sapir, after writing me that he was "fairly enthusias- 
tic" about it, published laudatory reviews in The Nation, The 
Freeman and The Dial; Laufer and Wissler expressed themselves 
orally in eulogistic terms; Elsie Clews Parsons hailed the book in 
The New Republic. John M. Cooper even now rarely sees me 
without making favorable comments on the work and urging me 

Autobiographical Data 

to bring it up to date. Malinowski was consistently laudatory in 
oral and published statements; even Goldenweiser went so far 
as to call it "a good book." Father Wilhelm Schmidt was eager 
to have it translated into German, a project thwarted only by the 
Austrian financial conditions at the time. On the other hand, 
Payot was willing to publish Mme E. Metraux's translation fifteen 
years after the original edition under the title of Traite de sociolo- 
gie primitive; and subsequently there was even a Japanese trans- 

From a sordid point of view, Primitive Society was the only 
book I ever wrote except for my elementary Introducton to Cul- 
tural Anthropology (1934; 1940) that proved profitable. Whereas 
some others failed to recoup me for typing expenses, six thousand 
copies of Primitive Society were sold in the United States; and as 
early as 1924 the royalties paid for a trip to Europe. 

Looking at the book objectively, I recognize more defects than 
the severer critics have brought out. My preparation for the task 
I set myself was wholly inadequate. Apart from the problems of 
associations and of the North American kinship and clan systems 
I had not been particularly interested in social organization. More- 
over, as a museum worker I had not had an opportunity to or- 
ganize my ideas on relevant questions except during a semester 
course as a visitor at the University of California. In consequence 
I had never considered many vital questions when I set pen to 
paper and was grossly ignorant even of what was then ascertain- 
able concerning African and Oceanian phenomena. Having rashly 
promised my then publisher to furnish the manuscript within 
four months, I very imperfectly filled the gaps in my knowledge 
during that span of time. 

Why, then, the general recognition by the crowd? Several 
factors have to be considered here. First, there had been no 
general treatise since Morgan's Ancient Society (1877). Since 
that date field research had accumulated vast stores of intrinsically 
interesting material, of which my book, however inadequately, 
gave the reader some conception. It afforded some insight into 
the phenomena of social units based on kinship, such as had been 
treated by Morgan, and also those first systematically dealt with 
by Schurtz. Experience seems to show that a reader is always 

Autobiographical Data 

grateful for his first orientation in a major field, and my book 
profited from the lack of any handbook on primitive sociology 
that was at all up-to-date in its standpoint. Secondly, Primitive 
Society embodied very largely not my original conceptions, but 
the anti-unilinear and diffusionist views that had gained mo- 
mentum in both Europe and America. Specifically, my criticism 
of older views fell in with the picture of early family life as con- 
ceived by Boas, Swanton, Malinowski, Schmidt, and others; I 
happened to be the first, not to voice them, but to voice them in 
a general work, thus accidentally appearing as the spokesman of 
an important body of anthropological thinkers. Thirdly, I natu- 
rally incorporated in the book the conclusions I had previously 
set forth concerning Plains Indian associations and relationship 
terms in technical papers. These had naturally had a limited 
public, thus whatever merit belonged to them was attached by 
most readers to the general book. 

One topic that, I believe, was really advanced in Primitive 
Society was that of property rights. That they existed on even 
very primitive levels was, in my opinion, established in the rele- 
vant chapter; which has been specially praised by such sociologists 
as Franklin H. Giddings and William F. Ogbum. Subsequently I 
slightly enlarged on one aspect of the subject in an article on 
"Incorporeal Property in Primitive Society" {Yale Law Journal, 
XXXVII [1928], 551-563).^ 

Primitive art and literature. — My observations on primitive art 
in general merely followed the lines traced by my predecessors, 
— Boas, Wissler, and Kroeber, though possibly in the study of 
Plains Indian rawhide bags I more than earlier students stressed 
the total configuration of the decoration rather than its constituent 
elements ( Crow Indian Art, AMNH, Anthropological Papers, XXI 
[1922], 27111.). However, I attach greater importance to exact 
measurements of comparable designs in neighboring tribes ("A 
Note on Aesthetics," American Anthropologist, XXIII [1921], 
170-174).^ Though wholly neglected except by Spier, this paper 
bears on the psychology of aesthetic perception and points the 
way toward a more refined comparative teclinique than is com- 

Reprinted as No. 18 in this volume. 
Reprinted as No. 10 in this volume. 


Autobiographical Data 

monly in vogue. It was stimulated by Gustav Theodor Fechner's 
speculations on the "golden section" and the ideal rectangle. 

Apart from routine recording of aboriginal tales, partly in the 
original, and indicating their historical connections, I have con- 
cerned myself with stylistic questions and the problem of the na- 
tive narrator's creative processes. Specifically, I have dealt with 
the degree of accuracy with which a traditional text is repro- 
duced ( Studies in Plains Indian Folklore, Univ. Calif. Publ. Amer. 
Arch, and Ethn., XL [1942], 1-28). 

Religion. — My conception of ceremonialism as largely a com- 
bination of spectacular non-religious elements secondarily as- 
sociated with the religious aspects ("Ceremonialism in North 
America," American Anthropologist, XVI [1914], 602-631) ^ was 
a logical extension of Boas's principle of secondary association, 
already exemplified in my doctoral dissertation with reference to 
mythology and later in the study of Plains societies. 

My general treatment of the entire subject in Primitive Re- 
ligion met with a curious variety of responses. One of the oddest 
features about them is a disparity on one point. Goldenweiser an- 
nounced in a review that I had merely dealt with the paraphernalia 
of religion, Sapir wrote me in corresponding fashion, Radin ex- 
pressed himself similarly to others, and Ogbum has recently 
pronounced it "singularly defective in giving the reader any 
feeling for the spirit of rehgion" (Wm. F. Ogbum and Meyer F. 
Nimkoff, in Sociology, Cambridge, 1940, pp. 69 £F.). Considering 
the uncanny concern of these gentlemen for the interests of re- 
ligion, I note the absence of this particular line of criticism among 
its ofiBcial representatives. Archbishop N. Soderblom, Father Wil- 
helm Schmidt, Monsignor J. M. Cooper. Professor Robert L. Cal- 
houn of the Yale Divinity School and one of his colleagues whose 
name has slipped my memory have orally expressed their satis- 
faction with the book. 

As to my "contributions" in this field, my own opinion is as fol- 
lows. So far as I know, I was the first ethnographer to attempt a 
tie-up between the ethnographic data and scientific psychology. 
Specifically, I emphasized the occurrence of individual varia- 
bility, and not merely with regard to visionary experiences. Fm*- 

® Reprinted as No. 24 in this volume. 


Autobiographical Data 

ther, I suggested several problems for future research, such as the 
role of women in primitive religion and the nature of the ideas 
that are associated by primitive man in his beliefs and cults. 

My position toward reHgion as a cultural phenomenon springs 
directly from my conception of anthropology as a science. A 
science of culture must take cognizance of values because these 
form an essential part of its subject-matter, but it must treat them 
objectively. The technique I use for doing this is verbatim quota- 
tion: The ineffable cannot be described by wordy ravings of an 
expositor, which must be "singularly defective in giving the reader 
any feehng for the spirit of religion." But the utterances of the 
religious person himself will come as close as anything can to 
suggesting what the thrill of ecstasy and other phenomena of faith 
are like; they are certainly more authentic than anything Golden- 
weiser or Benedict or anyone else might say on the subject. 

Further, an anthropological point of view must steer clear of 
judging the values thus set forth and must consistently maintain 
its benevolent neutrality not only in dealing with West African 
fetichism, but even in the face of Roman Catholicism or Greek 
Orthodoxy. Some of the most eminent anthropologists have failed 
to attain tliis point of view, and for myself it was not easy to rise 
above the encyclopaedist attitude of my early manhood. It was 
the second section of chapter iv in Mach's Die Mechanik (Leip- 
zig, 1912), pp. 429 ff., that roused me from my skeptical inertia 
and made me see the Church in proper perspective as against 
the force of traditional thinking quite apart from ecclesiastical 
doctrine. The substitution of a genial attitude toward contem- 
porary faiths for my earlier anticlericalism occurred between 
1918 and 1924; it has been a persistent part of my philosophy; 
and ever since, as in my popular book Are We Civilized? (New 
York, 1929), 1 have been fond of pointing out in the interests of 
fairness the irrationalism of non-clerical writers. From oral and 
printed statements of my elders and contemporaries I infer that 
in this position, which I regard as the only justifiable one scientifi- 
cally, I stand very nearly alone in these age-classes. 

For the meager appreciation of my book there were both in- 
trinsic and accidental reasons. For one thing, it did not suivey 
the entire field systematically, as Primitive Society in principle 


Autobiographical Data 

had surveyed its own. Secondly, I now appeared not as the spokes- 
man of a group, but as a free-lance reporting on his adventures 
into by-paths that had lured him. It was too much to expect that 
many readers, even if otherwise kindly disposed, should share 
my own predilections. Thus, Father Schmidt reviewed Primitive 
Religion favorably enough, but with reference to the problems 
that interested him, not me, in this connection; the parts I at- 
tached importance to, i.e., those concerned with psychology, he 
passed over with perfunctory commendation while lavishing 
much space on historical questions which I had explicitly de- 
clared as of minor significance from my point of view. According 
to less benevolent critics — Goldenweiser and his following, e.g. 
— I was getting above myself in daring to write on a subject too 
subtle for my pedestrian comprehension. Though I do not ac- 
cept this judgment as objectively valid, I am deeply sympathetic 
with it as a subjective biological reaction. Intuitively sensing my 
appraisal of their scientific worth, the critics naturally resented 
it; hence also my approach to scientific problems. In maturer 
years I found solace against this form of detraction in Goethe's 
reflections on Kotzebue; at a less serene stage I confess that my 
sentiments were nearer those of Goethe's Gotz von Berlichingen 
in a famous unquotable quotation. 

South American ethnography. — In 1926 the late Baron Erland 
Nordenskiold first called my attention to Curt Nimuendaju of 
Belem do Para as an exceptionally fine ethnographer, but I was 
then little interested in South America. Subsequently Norden- 
skiold's student. Dr. Itzikowitz, asked whether something could 
be done for Nimuendaju, who himself wrote me a letter and sent 
a sample manuscript. In 1935 I was able to get him a grant for 
field work from the Institute of Social Sciences of the University 
of California and to have this renewed for a number of years, 
sometimes eked out or superseded by grants of other institutions. 
In consequence Nimuendaju revisited various Ge tribes and, 
more recently, the Tukuna, presenting possibly the best all-round 
studies in existence on the simpler South American tribes. These 
I translated from his German manuscripts and, so far as possible, 
published in English translation after a copious correspondence 
conducted by both of us in German. The Apinaye (The Catholic 


Autobiographical Data 

University of America, Anthropological Series, No. 8, Washing- 
ton, D.C., 1939) and The Serente (Publications of the Frederick 
Webb Hodge Anniversary Publication Fund, Vol. IV, Los An- 
geles, 1942) are the most substantial publications up to date, 
but a far more comprehensive treatise, "The Eastern Timbira," 
which is to appear in the University of California series, has 
just reached the galley proof stage (February, 1944).^ A sizable 
Tukuna monograph remains untranslated.^ 

The extraordinary accuracy of Nimuendajii's observations, 
which revealed unexpected phenomena in the sociological and re- 
ligious culture of South American tribes, stimulated a number 
of brief papers setting forth their theoretical import for the re- 
construction of American culture history or for the problems of 
culture growth generally. The point I especially emphasized was 
the remarkable tendency of the simple Ge tribes to strike out 
along independent paths from an obviously common cultural base 
("A Note on the Northern Ge Tribes of Brazil," American An- 
thropologist, XLIII [1941], 188-196). 

My acquaintance with Nordenskiold led to another successful 
bit of promoting. When I was in Sweden in 1924, he and Rivet 
had told me of a Handbook of South American Indians which 
they had in mind as a joint venture. In 1932, when hard put to it 
to justify to my conscience my drawing a salary as Ghairman of 
the Division of Anthropology and Psychology of the National Re- 
search Council, I proposed that the Council foster a Handbook 
of the sort indicated and ask Nordenskiold to edit it. He was 
favorably inclined, but died soon after. For years nothing hap- 
pened except that the Division continued a committee concerned 
with the project and that the Smithsonian Institution assumed 
moral sponsorship. However, when Latin-American relations 
loomed on the political horizon, Congress was persuaded to make 
an appropriation and the State Department came to take a live 
interest in the work as tending to promote pan-American amity. 
Dr. Julian H. Steward of the Smithsonian, one of my former stu- 

'' Curt Nimuendaju, The Eastern Timbira, Univ. Calif. Publ. Amer. Arch, and 
Ethn., Vol. XLI (1946). 

* Translated by William D. Hohenthal and edited by Robert H. Lowie, The 
Tukuna was published in 1952 as Univ. Calif. Publ. Amer. Arch, and Ethn., Vol. 


Autobiographical Data 

dents and also of Nordenskiold's during his stay at Berkeley, 
became editor in chief and successfully rallied many contribu- 
tors. As a result the material for the Handbook has been ac- 
cumulating at a terrific rate. Printing on the first volume is to 
begin at the time of writing (February, 1944). 

Theoretical position. — My general attitude toward cultural 
anthropology has consistently been that it is a science which re- 
quires precisely the same logical and psychological processes as 
any of the supposedly more exact sciences. That our subject 
matter also permits artistic treatment I have never denied, but I 
have tried to keep apart objective findings and aesthetic inter- 
pretation. My attitude toward certain modern fashions in anthro- 
pology should be considered in this light. I do not necessarily 
deny all significance to them, but I do not care to take up with 
them until I see some likelihood for objective procedure. This 
holds especially for the attempted synthesis of psychoanalysis 
and anthropology. 

In the traditional domain of anthropology I have been a "mid- 
dle-of-the-road" man. That is, I have, like Tylor and Boas, ac- 
cepted both diffusion and independent development according 
to the evidence in particular cases. Much more than Boas I have 
maintained a faith — not in "laws," indeed — but in the possibility 
of determining fruitful correlations, especially in the study of 
social structure. 


PART I Kinship and 
Social Organization 

The eight papers included here form only a minor part of Lowie's 
extensive publication on kinship and social organization. They 
serve nevertheless to illustrate the kinds of problems that concerned 
him and the critical acumen he brought to them. They serve also 
to illustrate his ever-widening command of ethnographic informa- 
tion, his tenacious preoccupations, and his willingness to admit 
error when convinced by careful rescrutiny of his position. 

Social Organization 

Like the generation of thinkers that preceded ours, 
we are living in an age of revolt, but the object of our revolt is 
different from theirs. Our predecessors fought tradition as arrayed 
against reason. We have the task of exorcising the ghosts of tradi- 
tion raised in the name of reason herself. There is not only a folk- 
lore of popular belief, but also a folklore of philosophical and 
scientific system-mongers. Our present duty is to separate scien- 
tific fact from its envelope of scientific folklore. This duty has 
been recognized by workers in various fields. And so we have in 
philosophy James's protest against monistic mythology; in physics 
and chemistry Mach's protest against mechanistic mythology; in 
biology and anthropology a no less vigorous protest against evo- 
lutionary mythology. Monism, mechanism, evolution are doubt- 
less valuable concepts; but they are valuable in proportion as they 
are free from scientific folklore. 

Our present course of lectures is designed to help in the separa- 
tion of anthropological fact from anthropological folklore. This is 
the more necessary because not only laymen but even scientists 

Two lectures (in a course of four) delivered in January under the auspices of 
the Department of Anthropology, American Museum of Natural History. Printed in 
American Journal of Sociology, XX (July, 1914), 68-97. 


Kinship and Social Organization 

of neighboring fields — historians, economists, sociologists, social 
reformers — continue to use as definitely established truths an- 
thropological theories that are now gracing only the refuse heaps 
of the modern anthropologist's laboratory. 

In my own lectures, I will confine my attention to a single 
problem — that of the clan and the gens, or, to call both by a 
generic term, that of the one-sided exogamous kinship group, a 
group that traces descent either through the father or through the 
mother exclusively, and within which marriage is prohibited 
(exogamy). I will outline the conception of this group in older 
anthropological literature. I will show why we secessionists re- 
volt against that conception. And I will attempt to show how 
nowadays we grapple with the range of facts that concept was 
intended to summarize. 

The concept we are here concerned with has been most clearly 
defined by Lewis H. Morgan in his Ancient Society, a work that 
has molded the views of innumerable professional and non-pro- 
fessional students of anthropology on the social organization of 
primitive tribes. This is not the place to define accurately Mor- 
gan's place in the history of anthropology. To avoid misunder- 
standing, I will state at the outset that in my opinion that place 
will remain a high one. But we must distinguish between Morgan 
the observer, and Morgan the theorist; and in Morgan's theoreti- 
cal work we must again distinguish between his unusual power 
to see the importance of certain facts that had escaped others, 
and the very ordinary power shown in his naively synthetic con- 
structions. Morgan's observations have indeed been challenged 
in part, yet in almost every instance, not only with reference to 
the Iroquois but also as regards tribes he was less intimate with, 
they have been corroborated by later and more thorough investi- 
gation. We are, therefore, entitled to consider him a painstaking, 
trustworthy observer. On the other hand, Morgan's interpretation 
of human society as a whole was not only unduly colored by his 
personal observations among the Iroquois, but reflected the trend 
of his age toward artificial evolutionary schemes. To develop such 
a scheme requires more than average ability, but, contrary to 
current notions, it does not require a very high grade of ability, 


Social Organization 

certainly not of scientific ability. This Morgan displayed in a far 
more convincing manner when he noted the character of the 
Iroquois kinship system as distinct from our own, defined it, and 
set about with truly Darwinian industry to determine its ana- 
logues the world over. This genuinely scientific and theoretically 
important undertaking was doubtless not so spectacular as the 
interpretative speculations he superadded to the facts, but it will 
be rated higher by future generations. 

To attack our problem. In Ancient Society Morgan's general 
aim is to trace the history of social organization from the period 
of savagery to that of latter-day civilization. This development, 
he contends, took place through a series of unconscious reforma- 
tory movements enforced by natural selection. Low down in the 
scale of savagery there was a period of intermarriage of brothers 
and sisters in a group. At a later stage this was prevented by 
forming social units that would include brothers and sisters (as 
well as many other members of the tribe), and prohibiting mar- 
riage between all members of the new units. These organizations 
were of two distinct types according to whether kinship was 
traced through the mother or the father: they were either what 
are now known as "clans" or what we now call "gentes." A clan 
consists of "a supposed female ancestor and her children, to- 
gether with the children of her female descendants, through fe- 
males, in perpetuity." A gens consists of "a supposed male ancestor 
and his children, together with the children of his male descend- 
ants, through males, in perpetuity." Both the clan and the gens 
would bar intermarriage of brothers and sisters, and also marriage 
of cousins, no matter how distant, belonging to the same kinship 
group. On the other hand, marriage was not thereby prevented 
between all blood-relatives. With female descent, for example, I 
should not be permitted to marry my mother's sister's daughters, 
because they belong to my own clan, but I might marry my 
father's sister's daughters, who would necessarily belong to an- 
other clan. Morgan believes that, once invented, the scheme of 
the one-sided exogamous kin group spread "over immense areas 
through the superior powers of an improved stock thus created." 
With the exception of Polynesia, it formed "the nearly universal 


Kinship and Social Organization 

plan of government of ancient society, Asiatic, European, African, 
American, and Australian." ^ 

To this notion of the one-sided exogamous kin group Morgan 
added a theory of how that group developed from an archaic to 
a relatively modern form. In order to secure "the benefits of 
marrying out with unrelated persons," it would obviously be im- 
material whether kinship is reckoned on the mother's or father's 
side, so long as an equal number of relatives were prevented from 
mating. But Morgan holds that at the time when the one-sided 
exogamous group originated "marriage between single pairs was 
unknown, and descent through males could not be traced with 
certainty" (p. 67). Hence, he contends, in the archaic form of 
the kin group, kinship could be reckoned only in the maternal 
line, which also determined inheritance. When the paternity of 
children was assured, Morgan assumes that fathers revolted at 
their children being disinherited by the clan rule of inheritance, 
and in this way descent in the female line was overthrown and 
patrilineal descent substituted: in modern terminology, the clan 
was changed into a gens. Such, at least, is Morgan's account for 
the change among the Greeks and Romans (p. 345). For the In- 
dian tribes with gentes he does not venture to suggest tliroughout 
the same motive for the development of the gens from the clan. 
Speaking of the Siouan family, he writes: 

It is surprising that so many tribes of this stock should have 
changed descent from the female line to the male, because when 
first known the idea of property was substantially undeveloped, 
or but slightly beyond the germinating stage, and could hardly, 
as among the Greeks and Romans, have been the operative cause. 
It is probable that it occurred at a recent period under American 
and missionary influences (p. 157). 

In general it may fairly be said that Morgan regards descent 
traced through the father as a quite recent institution, and be- 
lieves in the ancient universality of the clan among North Ameri- 
can Indian tribes (e.g., p. 177). 

'L. H. Morgan, Ancient Society (New York, 1877), pp. 27, 63, 74, 377-379. 
I am using the term "gens" not in Morgan's sense but in that now common among 
American ethnologists. 


Social Organization 

It is my intention to test Morgan's theory by the Indian data on 
which it is primarily founded. We may begin our test with the two 
most tangible questions that develop in connection with Morgan's 
scheme. Was the one-sided exogamous kinship group really a 
universal institution among the natives of North America? And 
did the exogamous gentes found among them develop uniformly 
out of exogamous clans? These questions have been answered by 
Dr. Swanton,^ in the light of modern investigation, and while still 
later research has corrected his statement of the case in detail I 
find myself in full agreement with his general conclusions. 

In answering the first question it would not serve our purpose 
to enumerate the tribes that have exogamous kinship groups and 
set off against them the tribes that have not. For in this manner 
the real meaning of the facts would often be obscured through 
lack of weighting. For the question of the ancient universality of 
the exogamous kinship group is not equally significant whether 
the institution occurs among two quite unrelated tribes or among 
two tribes which, like the Hidatsa and Crow or some of the South- 
ern Siouan tribes, have only branched out from a common an- 
cestral tribe during the last four or five centuries. And obvi- 
ously the recent adoption of a clan or gentile system, which in a 
fair number of instances is demonstrably a result of borrowing 
from neighboring tribes, is of no importance from this particular 
point of view. Fortunately the essential facts can be expressed in 
a somewhat summary fashion, owing to the geographical con- 
tinuity of the tribes possessing the system in question. We find 
it, roughly speaking, in the greater part of the United States, east 
of the Mississippi, and some of the adjoining Canadian territory; 
among the Caddo and Southern Siouan tribes of the Plains, as 
well as among several of the Northwestern peoples in the same 
area; in New Mexico and Arizona; on the coast of British Co- 
lumbia and Alaska and in part of the Northwest coast hinterland. 
There are thus four fairly continuous areas within which the one- 
sided exogamous kin group is known to exist. In the remaining 
part of North America north of Mexico no such institution has 

^ John R. Swanton, "The Social Organization of American Tribes," American 
Anthropologist, VII (1905), 663-673. 


Kinship and Social Organization 

been discovered. Among the Eskimo, throughout the Mackenzie 
River and Plateau areas, as well as in nearly all of California 
and several of the Plains tribes, diligent inquiry has failed to re- 
veal any trace of such an exogamous system. It cannot be sup- 
posed that the system did exist in these regions but has escaped 
the notice of observers. For there is nothing esoteric about the 
rule that kin must not intermarry; and where exogamous groups 
occur the social activities connected with them are so prominent 
that, according to the experience of American field workers, 
their existence is very readily ascertained. The supporters of Mor- 
gan's views must, therefore, reckon with the fact that the sup- 
posedly universal organization simply does not exist in a very 
large part of North America. It might be asserted that the system 
had once existed everywhere, but that in certain districts it has 
disappeared. But this remains a baseless assertion in the absence 
of any proof that such a process has occurred and in the absence 
of any reason for such a process of degeneration in the regions 
concerned. It cannot be maintained either that the tribes in ques- 
tion have advanced beyond the clan or gentile stage. When we 
compare the culture of the Shoshone, Paiute, Thompson River 
Indians, and others lacking the one-sided exogamous kin group 
with the Pueblo Indians, Iroquois, Omaha, and others possessing 
the system, it is at once apparent that whether from the point 
of view of industrial arts, social life, or ceremonial activity the 
tribes possessing the system are the more advanced. Swanton 
has rightly emphasized the fact that almost all the tribes with a 
clan or gentile organization are agriculturists, while the rest are 
almost all non-agricultural. The weight of such considerations as 
these has led Frazer — in other ways a typical representative of 
the classical school in anthropology — to reject Morgan's position 
and to admit that the stage of the exogamous clan or gentile sys- 
tem had never been attained by "the more backward members of 
the Redskin family." ^ 

This statement of the facts must not, however, be interpreted 
to mean that the tribes in question represent a stage preceding 
that of the clan in Morgan's scheme. For in that case we should ex- 
pect no definite restriction of marriage, even between own broth- 

^J. G. Frazer, Totemism and Exogamy (London, 1910), III, 1-3. 


Social Organization 


ers and sisters,^ while on the contrary we almost uniformly do 
find additional restrictions based on consanguinity. To cite only 
a few random examples: Among the Central Eskimo, marriages 
of cousins, nephews and aunts, nieces and uncles are prohibited.^ 
Cousins are forbidden to marry among the Thompson River In- 
dians, and even second-cousin marriages are disapproved.^ In 
the Nez Perce tribe there were no restrictions of marriage except 
in the case of relatives, but even second or third cousins were not 
allowed to marry.''' In these cases, of course, a superadded one- 
sided exogamous kin system would not "secure the benefits of 
marrying out with unrelated persons," because these benefits are 
already secured by existing marriage restrictions based on bonds 
of consanguinity. Thus the tribes in question in no way fit into 
Morgan's scheme of social evolution. They are not more advanced 
than the tribes possessing exogamous kin groups, for their gen- 
eral culture is undoubtedly lower. But neither are they so low 
in their social customs as to require an exogamous kin system for 
the retrenching of consanguine marriages. If anything, we should 
have to say that in this particular point they are higher, that is, 
nearer to our own mode of conduct, than the tribes organized in 
exogamous kin groups as conceived by Morgan.^ It might still 
be argued that tribes may advance very unequally in different 
departments of culture; that therefore the loosely organized peo- 
ples may have lagged behind in their economic and industrial 
life while forging ahead of the tribes with clans or gentes in their 
social usages; that therefore they did once possess clans or gentes 
but have passed beyond that stage. The general principle on 
which such an argument would rest is sound, but its application 
is highly unconvincing in the present case. It would never be 
applied except to save the endangered hypothesis, involving as| 
it does an appeal not to any observable facts, but to our ignor-f 
ance of unobservable ones. Accordingly, we may dismiss it and 

* Except so far as the Australian four-class system prevailed, which, however, 
did not prevent first-cousin marriage. Morgan, op. cit., pp. 425, 503. 

" Franz Boas, "The Central Eskimo," BAE, 6th Ann. Rept. ( 1888), p. 579. 
^ James Teit, The Thompson Indians of British Columbia, AMNH, Memoirs, I 
(1900), 325. 

' H. J. Spinden, The Nez Perce Indians, AAA, Memoirs, II, Part III ( 1908), 250. 

* This conception of Morgan's will, however, prove to be erroneous. 


Kinship and Social Organization 

sum up our conclusion to the effect that in North America ex- 
ogamous kin groups, instead of being universal, were absent from 
a great many tribes, and that these for the most part possessed 
a less complex culture than those who had this institution. 

Let us now turn to our second problem. Has the history of the 
one-sided exogamous group in North America been the origin 
and partial persistence of the archaic clan and its partial trans- 
formation into a gens? In other words, have all the gentes found 
among our Indians been preceded by a clan system? It will be 
best to consider first the mechanism by which Morgan conceives 
the change to have occurred and then the question of fact in- 

Roughly speaking, we find the gentile ( as opposed to the clan ) 
system among the Central Algonkian, Blackfoot, and Southern 
Siouan Indians. For the last-mentioned tribes Morgan suggests, 
as already noted, that the hypothetical change occurred under 
American and missionary influences. This suggestion, however, 
is anything but convincing. We know of a number of instances 
where civilization has introduced novel social arrangements 
among Indian tribes, but of none where it has produced Morgan's 
hypothetical development. Among the Crow, for example, the 
government has introduced a patrilineal rule of property inher- 
itance, but the native rule of maternal descent continues to hold 
for clan names and afiBliations. The Iroquois have adopted the 
system of passing on surnames given by whites from father to 
son, but the ancient matrilineal system remains in full force. The 
Crow have probably been subject to wliite influence for as long 
a period as the Southern Siouan tribes, and the Iroquois doubt- 
less for a much longer period. It seems highly improbable that 
within the short period of something like a century contact with 
civilization should have caused a considerable number of tribes 
not merely to adopt the wliite way of reckoning descent in mat- 
ters that would be of moment in their dealings with whites, but 
to be so thoroughly imbued with the point of view of the whites 
as to adopt the alien mode of tracing lineage in aU parts of their 
social system. The Mandan, of whom not a dozen full-blood mem- 
bers survive at the time of writing, still reckon tribal affiliation 
according to the matrihneal scheme; cliildren of Mandan mothers 


Social Organization 

and Hidatsa fathers are Mandan; children of Hidatsa mothers are 
Hidatsa. Morgan's suggestion as to the cause of change of de- 
scent among Siouan tribes may therefore be dismissed as unsatis- 

With reference to the Algonkian Shawnee, Morgan makes a 
suggestion more in accord with his general scheme of develop- 
ment (p. 169). Instead of ascribing the change of descent to 
civilized influence, he is here inclined to assign an internal cause 
— the wish to enable a son to succeed his father as chief, and to 
enable children to inherit property from their father. But, re- 
peating in essence the foregoing remarks, we must insist that both 
these questions — descent of office and descent of property — do 
not necessarily affect the fundamental matter of reckoning lin- 
eage. The Crow illustration cited above fits in here also, for it is 
manifestly a matter of indifference whether the rule of inherit- 
ance is changed from alien or indigenous causes. The question is 
whether a change in the rules of property inheritance from the 
maternal to the paternal line is itself a cause of changing clan 
affiliations into gentile affiliations; and there seems to be no evi- 
dence for this alleged causal connection. 

This does not answer the question of fact whether, regardless 
of what causes may have operated, the gens is a development 
from the clan. Morgan's proof consists essentially in pointing out 
that while certain tribes have a gentile system other members of 
the same stock have clans. This is of course a two-edged argu- 
ment that may with equal force be used to prove that clans de- 
veloped from gentes. From the fact that Mandan, Hidatsa, and 
Crow reckon descent in the female line, Morgan argues that the 
Ponka, Omaha, Iowa, and Kaw formerly reckoned descent in the 
same way (pp. 155 ff.), all these tribes speaking Siouan lan- 
guages. So, from the occurrence of female descent among the 
Delaware, Morgan infers "its ancient universality in this form in 
the Algonkian tribes" (p. 172). To be sure, this conclusion is sup- 
ported by some additional data. The Delaware are declared to 
be "recognized by all Algonkian tribes as one of the oldest of 
their lineage," though it is safe to say that many Algonkian tribes 
were blissfully ignorant of the very existence of the Delaware in 
Morgan's time. Morgan furnishes better evidence in citing cases 


Kinship and Social Organization 

of several Algonkian tribes with male descent where nevertheless 
the chief's office was passed, not from father to son, but from 
maternal uncle to sister's son (pp. 166, 170). However, these 
cases are very few, have not been corroborated by later inquiry, 
and admit of other explanations. For example, there may be spe- 
cial rules for the inheritance of certain oflBces distinct from those 
which otherwise hold. The coexistence of diflFerent rules of de- 
scent for different social groups is well estabhshed in various 
primitive tribes. Thus, in Uganda descent of clan membership 
was patrilineal for all except princes of royal blood, who were 
always reckoned of kin with their mother. Considering that even 
with the most favorable interpretation of the cases cited by Mor- 
gan we are stiU confronted with a considerable number of tribes 
with paternal descent and no trace of any other system, we must 
conclude that Morgan has not established his scheme of develop- 
ment inductively but deduced it from his a priori postulate of un- 
knowable fatherhood in archaic times. 

This brings us face to face with a most important theoretical 
problem. We have indeed shown that Morgan has not proved his 
case from the North American data; but he may nevertheless be 
right if others have established the general law that matrilineal 
descent precedes paternal descent. Extending our inquiry beyond 
the American data, we must adinit that until recently most so- 
ciologists and anthropologists deduced this sequence from such 
postulates as the uncertainty of fatherhood among primitive con- 
ditions. Tylor's point of view was doubtless in large measure de- 
termined by such considerations, but he supports it on a more 
solid basis of fact than is usually the case, and accordingly it will 
be best to consider his reasoning in some detail. Advancing what 
he himself characterizes as a geological argument, he holds that 

. . . the institutions of man are as distinctly stratified as the earth 
on which he hves. They succeed each other in series substantially 
uniform over the globe, independent of what seem the comparatively 
superficial differences of race and language, but shaped by similar 
human nature acting through successively changed conditions in 
savage, barbaric, and civilized life. 

Tylor groups primitive tribes under three headings, con'e- 
sponding to successive cultural strata: those with a maternal sys- 



i- \ 

Social Organization 

tern of descent, those in which both maternal and paternal rules 
of descent coexist, and those with a purely paternal descent. He 
then examines, with reference to their occurrence in these strata, 
certain social customs — notably the remarriage of widows and 
the "couvade." His treatment of the latter case will sufiBce to il- 
lustrate the method of reasoning followed. The couvade is the 
practice (found most conspicuously in some parts of South Amer- 
ica ) by which "the father, on the birth of his child, makes a cere- 
monial pretense of being the mother, being nursed and taken 
care of, and performing other rites, such as fasting and abstain- 
ing from certain kinds of food or occupation, lest the new-born 
should suffer thereby." Tylor finds not a single instance of this 
strange usage among purely maternal peoples. In the maternal- 
paternal condition there are not less than twenty eases, while in 
the paternal the number dwindles to eight. From this Tylor infers 
that the purely maternal stage is the 'earliesf'because there is no ^ 
survival of the couvade from other stages as there is in paternal 

Just as the forms of life, and even the actual fossils of the Car- 
boniferous formation, may be traced on into the Permian, but 
Permian types and fossils are absent from the Carboniferous strata 
formed before they came into existence, so here widow-inheritance 
and couvade, which, if the maternal system had been later than the 
paternal, would have lasted on into it, prove by their absence the 
priority of the maternal.^ 

In support of Tylor's theory, that matrilineal institutions pre- 
cede patrilineal descent, concrete evidence of all kinds has been 
adduced. Among the most recent writers. Rivers has expressed 
the conviction that this sequence holds for Oceania. ^° On the other 
side, American ethnologists have appealed to the case of the 
Kwakiutl of British Columbia, where there is assumed to have 
taken place a change in the contrary direction. According to 
Professor Boas, the Kwakiutl, like the tribes of Oregon, Washing- 
ton, and southern Vancouver Island, once lived in village com- 

"E. B. Tylor, "On a Method of Investigating the Development of Institutions, 
Applied to Laws of Marriage and Descent," RAI, Journal, XVIII ( 1889), 245-269. 

" W. H. R. Rivers, "Survival in Sociology," Sociological Review, VI (1913), 


Kinship and Social Organization 

munities with paternal descent. Owing to the influence of the 
more northern Pacific tribes, whose system is matriUneal, the 
Kwakiutl grafted the northern principle of descent on that of 
the south, with the result that certain privileges are inherited in 
the paternal line and a much larger number are obtained by mar- 
riage tlii'ough an intricate method that insures maternal descent. ^^ 
But although the Kwakiutl facts are very interesting, it is highly 
doubtful whether they have the theoretical significance ascribed 
to them. It is, in the first place, worth noting that they represent, 
in Tylor's terminology, not a maternal but a maternal-paternal 
stage. At best, therefore, they yield evidence of change from a 
purely paternal to a mixed condition. Secondly, maternal descent, 
so far as it prevails, seems to be restricted to the inheritance of 
property, while the reckoning of a child's afiBliation seems to be 
indeterminate, as we have been more recently informed by Boas 
that a child is reckoned as belonging to both his father's and his 
mother's family. Thirdly, it is a matter of grave doubt whether 
the Kwakiutl units of which maternal-paternal descent may be 
predicated correspond to the type of unit to which Morgan, at 
all events, applies the sequence advocated by himself and Tylor. 
For Morgan is speaking all the time of exogamous units, whether 
clans or gentes, and among the Kwakiutl there seems to be no 
definite rule of exogamy but only a preference for marriage out 
of the group, and even this is denied in a later statement. ^^ Fi- 
nally, the Kwakiutl conditions are so specialized that adherents 
of the Tylor- Morgan theory may well regard them as exceptional; 
and even if the change from paternal to maternal descent be ad- 
mitted, it is possible to suppose a pristine stage of matrilineal 
reckoning preceding the patrilineal village communities. 

For these reasons the Kwakiutl conditions do not seem to fur- 
nish a favorable test case. Nevertheless, they embody the prin- 
ciple that forms the most vital objection to the classical theory as 
to rules of descent. For the Kwakiutl have developed their sys- 
tem not solely through internal growth but through contact with 

" Franz Boas, "The Social Organization and the Secret Societies of the Kwakiutl 
Indians," USNM, Report for 1895, pp. 334-335. 

^ Idem, "Tribes of the North Pacific Coast," Annual Archaeological Report, 
1905 (Toronto, 1906), pp. 239-240; A. A. Goldenweiser, "Totemism, an Analytical 
Study," Journal of American Folk-Lore, XXIII (April-June, 1910), 187, 213. 



Social Organization 

other tribes. The far-reaching influence of such connection with 
neighboring tribes generally has been realized to an increasing 
degree by modem anthropologists, and it obviously interferes 
with the doctrine of parallelism advanced by Tylor. For, grant- 
ing that on account of the similarity of human nature, human in- 
stitutions tend to succeed one another "in series substantially 
uniform over the globe," the borrowing of institutions would in 
an indefinite number of cases produce an abnormal sequence. 
We cannot even assert that where the observed sequence cor- 
responds to the theory the result is due to uniform causes pro- 
ducing parallel evolution. Among the Carrier and Babine Indians 
there is matrilineal descent. As the majority of the Northern 
Athapascans, of whom these tribes form part, have a loose organi- 
zation, it may be safely assumed that the Carrier and Babine 
once shared this sociological characteristic, provided we can in- 
dicate the conditions that in their case produced a change. We 
thus seem to have an illustration of the evolution of a clan system 
from the "earlier and less organized and regulated condition" 
postulated by both Morgan and Tylor. But the conditions that 
produced the change were not so general as the psychological 
constitution shared by humanity, but lay in the geographical con- 
tiguity of the Northwest Coast Indians, whose social organiza- 
tion was simply copied by the tribes in question. Accurate in- 
formation as to the actual process of cultural development has 
largely shattered the belief once held in the necessity of parallel 
evolution among unrelated tribes. Many ethnologists now hold 
that historical processes are unique in character, that every phase 
of human history is so complicated by individual traits that no 
laws of historical development can be framed. This view has so 
deeply affected modern anthropology that even in quarters pe- 
culiarly liable to classical influence a far more cautious formula- 
tion is now in vogue. It is no longer contended that every gentile 
system has superseded a clan system, but merely that if the rule 
of descent changes at all, it changes from matrilineal to patrilineal 
descent. Thus, N. W. Thomas writes: 

. . . whereas evidences of the passage from female to male reckon- 
ing may be observed, there is virtually none of a change in the op- 
posite direction. In other words, where kinship is reckoned in the 


Kinship and Social Organization 

female line, there is no ground for supposing that it was ever 
hereditary in any other way. On the other hand, where kinship is 
reckoned in the male line, it is frequently not only legitimate but 
necessary to conclude that it has succeeded a system of female 
kinship. But this clearly does not mean that female descent has in 
all cases preceded the reckoning of kinship through males. Patri- 
lineal descent may have been directly evolved without the inter- 
mediate stage of reckoning through females. ^^ 

And expressing a still more acceptable view, Cunow writes: 

Die meisten der heutigen vaterrechtlichen Halbkulturvolker 
haben sicherlich einst, wie sich deutlich aus ihren Rechtsbrauchen 
und Verwandtschaftsbezeichnungen nachweisen lasst, das Mutter- 
recht gekannt; aber das besagt noch nicht, wie den Vertretem der 
zweiten von Miiller-Lyer genannten Theorie eingeraumt werden 
muss, dass das Mutterrecht cine Institution ist, die sich bei alien 
Rassen und Volkem ohne Unterschied auf gewisser Entwicklung- 
shohe einstellt. Unter besonderen Umstanden mag das Mutterrecht 
ganz gefehlt haben oder doch die Mutterrechtsperiode von relativ 
kurzer Zeitdauer gewesen sein.^* 

From this modern point of view there is thus no reason to sup- 
pose that the Blackfoot, Central Algonkian, and Southern Siouan 
tribes ever possessed a clan system preceding their present or 
recent gentile system. Their general cultural condition, whatever 
may be the value of such a comparison, does not show a higher 
stage than that of maternally organized tribes; of the latter, in- 
deed, the Pueblo Indians are manifestly superior to any of the 
patrilineal tribes. Considering the modified form in which such 
sane students as Marett, Cunow, and Thomas now present the 
classical theory of father-right and mother-right, we may safely 
say that there is no reason why the patrilineal tribes of North 
America could not have developed their system directly from a 
loose organization without passing through the hypothetical in- 
termediate stage. Summing up, therefore, our reply to the two 
questions set at the beginning, we may say: 

1. It is as certain as anything can be from the nature of the 
case that the one-sided exogamous kin-group system, whether 

" N. W. Thomas, Kinship Organization and Group Marriage in Australia ( Cam- 
bridge: University Press, 1906), p. 15. Cf. R. R. Marett, Anthropology (New York, 
1912), p. 169. 

" Heinrich Cunow, in Zur Urgeschichte der Ehe und Familie, pp. 38-39. 


Social Organization 

in the forni of clans or gentes, was not universal among North 
American tribes. 

2. It is entirely unproved that those Indian tribes possessing 
a gentile system previously had a clan system, 


In the preceding discussion little has been said of what is really 
more important in Morgan's treatment than either the geographi- 
cal distribution of one-sided kin groups or the relative priority of 
different forms of these groups — Morgan's notion of what the kin 
group really represents. There can be no doubt that in Morgan's 
mind it was primarily not only a marriage-regulating agency (pp. 
74, 378), but the marriage-regulating agency, preceding a pro- 
hibition of marriage between blood-relatives generally. It was, 
moreover, the only original type of social unit he recognized in 
primitive society beyond a certain stage and before the develop- 
ment of political society (p. 63). Finally, it was everywhere the 
same (ibid.). 

Every one of these points is open to criticism. I shall first en- 
deavor to show that blood-relationship operated as a bar to mar- 
riage independently of the origin of the one-sided kin system. 

If Morgan's view of the kin group were correct, tribes pos- 
sessing this institution ought, in the first place, to consider mar- 
riage with the most distant cousin belonging to one's own group 
as incestuous as marriage with an own sister, for the exogamous 
rule according to hypothesis was a bar to brother-sister marriages 
only indirectly, inasmuch as brothers and sisters were included 
in the number of kinsfolk among whom marriage was tabooed. 
It is true that there are cases where union with a fellow-member 
of the same kin group is regarded as incest, even where there is 
no trace of blood-relationship between the mates. Among the 
Haida, for example, "so close was relationship held to be between 
persons of the same clan, that marriage within it was viewed by 
them almost as incest is by us." ^° But in other cases there is no 
such feeling. The Crow and Hidatsa have presei'ved their matri- 
lineal system to the present day, the former still considering 

^® J. R. Swanton, Contribution to the Ethnology of the Haida, Publications of tlie 
Jesup North Pacific Expedition (New York, 1909), I, 62. 


Kinship and Social Organization 

their clans exogamous. Yet transgression of the exogamous rule 
is a matter for ridicule not for punishment, is considered improper 
rather than criminal, and there is no evidence that any other point 
of view was ever taken by the natives. ^^ The same, according to 
Dr. Goldenweiser, applies to Morgan's own Iroquois: mild re- 
proval, not abhorrence, was meted out to the offenders. In these 
tribes marriage with an unrelated clansman would not begin to 
compare in offensiveness with marriage to a clansman who was 
a near blood-relative. 

It may of course be urged that the diflFerentiation of clansman 
and blood-relative is recent, but the argument may be supple- 
mented by another. Morgan's theory presupposes not only that all 
members of a kin group are equally barred from marriage, but also 
that marriage with members of other kin groups is permitted re- 
gardless of ties of consanguinity. This corollary is demonstrably 
false for the present time in a great number of cases, and with it 
falls the theory from which it is deduced. Among the Crow, mar- 
riage with a near relative on the father's side is as strictly tabooed 
as marriage with a near relative within the clan. The same ap- 
plies to the maternally organized Creek,^^ to the Navaho, Iro- 
quois, and Zufii.^^ Among the Omaha, while a man must marry 
outside of his gens, he is also required to marry outside of his 
mother's gens.^^ The same, to choose an example from an entirely 
different area, applies to the patrilineally organized Baganda of 
East Africa. ^"^ Among the Australians of Queensland blood-cousins 
are not allowed to marry though they belong to otherwise inter- 
marriageable groups. ^^ 

Against this twofold argument the only possible answer is 
that the kin-group affiliation was in each case primary; that the 
attitude toward marrying a related clansman, as compared with 

" R. H. Lowie, The Social Life of the Crow Indians, AMNH, Anthropological 
Papers, IX (1912), 188 f. 

" J. R. Swanton, "A Foreword on the Social Organizations of the Creek Indians," 
American Anthropologist, XIV (1912), 596. 

" Idem, "The Social Organization of the American Tribes," American Anthro- 
pologist, VII (1905), 667-668. 

"J. O. Dorsey, "Omaha Sociology," BAE, 3d Ann. Kept. (1884), p. 257. 

^ John Roscoe, The Baganda (London, 1911), p. 128. 

^^. E. Roth, Ethnological Studies among the NoHh-West-Central Queensland 
Aborigines (Brisbane, 1897), p. 69. 


Social Organization 

that toward marriage with an unrelated one, and the objection 
against marriage with related members of other clans developed 
after the feeling against kin intermarriage. Here it may be noted 
in the first place that absolutely no evidence exists for this view; 
it is pure dogma. On the other hand, there is good evidence for 
the reverse order of development. As noted above, it is precisely 
among those North American Indians possessing the simplest 
culture that ties of blood-relationship, and not of one-sided kin- 
ship, act as a bar to marriage. Examples could be drawn from 
among tribes of other continents. Thus, among the Kai of New 
Guinea, marriage between a brother and sister is forbidden, 
though nothing is said of any kin restriction. In the same area the 
Jabim prohibit unions between children of a brother and sister.^^ 
There is even more direct evidence. In several instances it ap- 
pears that the exogamic character of the clan or gens is derived 
from the feeling that all fellow-clansmen or gentiles are related 
by blood. Among the Blackfoot, 

when a proposal for marriage has been made, the relatives of the girl 
get together and have a talk, their first and chief concern being 
the question of blood relationship. Naturally, the band [gentile] 
affiliations of the contracting parties cannot be taken as a criterion, 
since both may have very near relatives in several bands and cousins 
of the first degree are ineligible. Should the contracting parties 
belong to the same band but be otherwise eligible, the marriage 
would be confirmed, though with some reluctance, because there 
is always a suspicion that some close blood relationship may have 
been overlooked. Thus, while this attitude is not quite consistent, 
it implies that the fundamental bar to marriage is relation by blood, 
or true descent, and that common membership in a band [gens] is 
socially undesirable rather than prohibitive. ... In any event, the 
attitude of the Blackfoot themselves seems to imply that the band 
[gentile] system came into existence after the present marriage 
customs and adapted itself to them rather than they to it.^^ 

A still more striking case is furnished by the Todas of Southern 
India. The Todas, according to Rivers, have a general term, piiliol, 

^ Keysser and Zahn, in Richard Neuhauss, Deutsch Neu-Guinea (Berlin, 1911), 
III, 89, 299. 

^ Clark Wissler, The Social Life of the Blackfoot Indians, AMNH, Anthro- 
pological Papers, VII (19II), 19-20; also in Current Anthropological Literature, 
I (1912), 15-16. 


Kinship and Social Organization 

for those relatives whose intermarriage is prohibited, including 
some that belong to the same gens and some who do not. A man 
casting about for a mate classified women as either piiliol or not, 
"and it seemed to me in several cases as if it came almost as a new 
idea to some of the Todas that his piiliol included all the people 
of his own clan [gens]." A Toda 

has not two kinds of prohibited affinity, one depending on clan 
[gentile] relations, and another on relations of blood-kinship, but 
he has only one kind of prohibited affinity, to which he gives the 
general term piiliol, including certain kin through the father and 
certain kin through the mother, and there is no evidence that he 
considers the bond of kinship in one case as different from the other 
as regards restriction on marriage. 

The fact that the Toda includes all those kin whom he may not 
marry under one general term, and that the kin in question include 
members both of his own and other clans [gentes], goes to show that 
the Todas recognize the blood-kinship as the restrictive agency 
rather than the bond produced by membership of the same clan 
[gens]. 24 , 

That is to say, among both Blackfoot and Todas such gentile 
exogamy as occurs is not a primary but a derivative phenomenon 
— is merely a function of the primary phenomenon, to wit, of re- 
striction on the basis of blood-relationship. 

To sum up. Restrictions of marriage based on blood-relation- 
ship apart from one-sided kinship coexist with kin-group exogamv. 
Restrictions of marriage due to blood-relationship exist where kin 
exogamy does not occur, that is to say, in North America at least, 
among the tribes with the most primitive culture. Finally, there 
is positive evidence that in some cases kin exogamy is a corollary 
of restrictions due to blood-relationship. The conclusion seems 
warranted that clan or gentile exogamy is a phenomenon super- 
imposed on prohibitions of incest in our sense. 

Let us now turn to the question whether the clan or gens is 
really the prototype of all social units in North American Indian 
society. Morgan found that among the Seneca-Iroquois the eight 
clans of the tribes were grouped in two divisions of four each. 
These larger divisions or "phratries," he argued, were nothing but 
overgrown clans that had become subdivided, the subdivisions 

^W. H. R. Rivers, The Todas (London, 1906), pp. 509-510. 


Social Organization 

preserving a consciousness of their kinship.^^ As association of 
clans occurs without a dual grouping I propose to call any such 
association a "phratry," and will call "moiety" one of two com- 
plementary divisions of a tribe, regardless of its relations to what- 
ever lesser divisions may coexist with the moieties. In my ter- 
minology the question then arises: Did the Seneca moieties 
develop from the Seneca clans by a process of subdivision? 

In order to solve the problem, let us consider the facts as pre- 
sented by Morgan and Dr. Goldenweiser. The Bear, Wolf, Beaver, 
and Turtle clans form one moiety; the Deer, Snipe, Heron, and 
Hawk the complementary moiety. While the clans bear animal 
names, the moieties are nameless. At the present, and indeed this 
even applied to Morgan's time, the moieties are not exogamous, 
but there is evidence that they were exogamous long ago; the 
clans remain exogamous even today. Each clan has a set of individ- 
ual names distinctive of its members, and there is a rule that none 
of these names shall be borne by more than one person at a time. 
The moieties do not possess distinctive sets of names, but on 
their part exercise certain functions not shared by the clans. At 
a ball game the division of players and of their supporters in bet- 
ting was by moieties; at tribal councils the chiefs of the two moi- 
eties sat on opposite sides, and voting on such a subject as the con- 
firmation of a new chief was by moieties; at a funeral the moiety 
complementary to that of the deceased conducted the cere- 
monies; finally, there were obscure religious functions connected 
with the phratries. 

These being the essential facts, what do we learn from them 
respecting the segmentation of two primary clans into eight? 
Before answering this question, let us try to determine whether 
we have any evidence for the segmentation of social units. Such 
evidence undoubtedly exists. Among the Onondaga-Iroquois, there 
is not only a Big Snipe but also a Little Snipe clan. If these were 
primarily distinct social units, we should expect to find that their 
sets of individual names were distinct, this being a characteristic 
of the clans of all the confederated Iroquois tribes. But the 
clans in question share the same set of names, whence their 
essential unity may safely be inferred. In other tribes additional 

^ Morgan, op. cit., pp. 88, 89, 91, 99. 


Kinship and Social Organization 

examples of segmentation could doubtless be adduced. Where 
diflFerent clans are named for animals of the same genus but for 
different species of that genus, there is at least some presumptive 
evidence for segmentation (though never more), and where the 
moiety bears the name of a predominant clan or gens within that 
moiety the same applies, though here the actual demonstrative 
value of the evidence is even less. 

However, in the case of the Iroquois it is not at all clear why 
the moiety and the clan should be supposed to be genetically re- 
lated. Either the moiety was not formerly exogamous — then there 
is no functional similarity between moiety and clan at all; or, 
as according to Morgan and Dr. Goldenweiser it is reasonably 
certain, the moiety was once exogamous. Then, also we are not 
justified in saying that both units were exogamous. For, at the 
time the moiety was exogamous, the clan was by logical neces- 
sity exogamous as a part of the exogamous moiety, while its dis- 
tinctively clan characteristics may have had nothing to do with 
exogamy .^^ The fact that when the exogamous rule of the moiety 
broke down it was limited to the clan proves nothing as to the 
unity of the two organizations. For in Australia, as Cunow shows, 
the marriage-regulating functions of the classes have in some in- 
stances been transferred to the genetically quite diflFerent totem 
kin units. It is true that an indefinite number of excuses can be 
given why proof of connection between the Iroquois moiety and 
clan should be lacking. Names of social units have been known 
to disappear; social units have been known to assume new func- 
tions; if the moiety developed according to Morgan's scheme, 
evidence of the exogamic character of the new clans must have 
disappeared. All such explanations remind one of the evolu- 
tionist's plea as to imperfections of the paleontological record. 
Such a plea is admissible where there is extraneous positive evi- 
dence, but does not fill the place of lacking evidence. In the 
case of evolution there is fortunately independent evidence; in 
our Iroquois case there is not. Hence, the unity of the Iroquois 
clan and moiety remains unproved, though not disproved. 

^^ This argument has been used by Dr. Goldenweiser with relation to Australian 


Social Organization 

The value of the foregoing discussion hes in several directions. 
For one thing it changes our view of the essence of the units com- 
monly called clans and gentes. Whenever such units form part of 
larger exogamous units, we can no longer assume dogmatically 
that they, too, are at bottom exogamous. This, nevertheless, re- 
mains a possibility, and on that assumption we must change our 
conception of the moiety. The moiety, instead of being a sub- 
divided exogamous clan, may be an association of exogamous 
clans constituting a unit of novel character even if it assumes 
the exogamous character once distinctive of the separate clans. 
And this again opens our eyes to all kinds of possibilities. We 
need no longer tug at all the facts of social organization in a vain 
ejBFort to thrust them into the strait-jacket of "exogamy." We be- 
gin to suspect that various types of social units may peaceably 
coexist in the same tribe, some regulating marriage, some, other 
social activities; nay, some regulating marriage in one sense, 
others, in a diflFerent sense. In short, instead of the dull uniformity 
of the theorists, we may have all the motley variety of real life 
with its profusion of individual differences. To justify this plu- 
ralistic view, let us turn to some facts. 

The Fox and Kickapoo are divided into exogamic gentes bear- 
ing animal names. So far these units, except for the rule of descent, 
correspond to the Iroquois clans. But in addition to their gentes 
both Fox and Kickapoo have a division into moieties that is utterly 
different from the Iroquois scheme. The Kickapoo child enters a 
moiety only after receiving a name. 

The name comes from the father's name, unless the right of nam- 
ing the child is handed over to the mother by the father. If the father 
is uskaca, then the oflFspring will be uskaca. If the mother is kicko'a 
and she has the right of giving the name, then the child is kicko'a. 
Again, the child can become a kickoa if he is given to a grandmother, 
grandfather, sister's son, or a sister's daughter; the child gets his 
name from the one in whose hands he falls, and if the namer is a 
kicko'a the child will be a kickoa. 

Among the Fox the father usually, but not always, determines 
which division a child shall enter. "If he is a To'kan, it is likely 
his children will be the same. Often the firstborn is the same as 


Kinship and Social Organization 

the father, and the next child is the other. No distinction is made 
on account of sex." "^ 

From an earHer account of the closely related Sauk it ap- 
pears that consecutive children of the same father were placed 
into different moieties, the oldest into his father's, the next oldest 
into the complementary moiety, and so forth.^^ The Fox and 
Kickapoo moieties are distinguished by the use of different paint 
for personal decoration. Their sole function has been limited to 
that of divisions in athletic games, but according to Dr. Michel- 
son, the Fox moieties had more serious (in part, ceremonial) 
duties. These moieties differ markedly from those of the Iroquois, 
not so much because of a difference in function, as in essential 
constitution. Functions may be assumed and lost. The moieties of 
the Fox might be supposed to have lost some of the characteristics 
found among the Iroquois, or the Iroquois moieties might be sup- 
posed to have originated as divisions similar to the Fox moieties 
and to have assumed additional duties. But the difference is 
more fundamental. While among the Iroquois all the members 
of several gentes are united in one moiety, this is contrary to the 
Fox scheme, where members of the same gens belong to opposite 
moieties, while either moiety probably joins together members 
of all the gentes. The problem that, however solved or however 
insoluble, naturally arises from the Iroquois facts, viz., whether 
the moiety is a subdivided clan or an association of originally dis- 
tinct clans, does not arise at all in connection with the Sauk and 
Fox. It would not even arise if the rule of descent were definitely 
patrilineal for the moiety as it is for the gens. For there is nothing 
to show that a man of gens a must belong to moiety A; and so 
long as there is no such definite correlation, the children of a given 
man, Aa, will indeed be also Aa, but will be joined for 
athletic purposes (or what not) by the children of An, who are 
also An. Consider Germans divided into Catholics (A) and Protes- 

-' William Jones, "Notes on the Fox Indians," Journal of American Folk-Lore, 
XXIV (1911), 216, 220; idem, "Kickapoo Ethnological Notes," American Anthro- 
pologist, XV (1913), 335. According to Dr. Michelson the firstborn child belongs 
to the moiety complementary to its father's, the second to its father's moiety, and 
so forth. 

"■* Thomas L. McKenney and James Hall, History of the Indian Tribes of North 
America (Edinburgh, 1933), I, 117. 


Social Organization 

tants ( B ) . From the point of view of marriage the German Catho- 
Kcs form as definite an endogamous unit as the Fox gens forms 
an exogamous unit. But Germans are also divided into socio- 
economic groups, a to n, within which membership is in a meas- 
ure hereditary, and when we are told that a German is of the 
landed gentry, a capitalist, a proletarian, etc., we do not know 
whether he is a Catholic or a Protestant, nor will it occur to us 
to ask, whether the religious division grew out of the economic 
division or vice versa. 

An even more instructive case is furnished by the Yuchi, 
formerly resident in Georgia and Alabama but now settled in 
Oklahoma. Like the Fox, the Yuchi are divided into exogamous 
groups, though with them descent in these groups is matrilineal. 
Like the Fox again, the Yuchi are divided into moieties, but with 
the important difference that membership descends in the pa- 
ternal line. Here again each moiety will thus embrace members of 
all possible clans and each clan will have members of both moi- 
eties. The functions of the Yuchi moieties are of great importance. 
Not only does the division during ball games follow moiety lines, 
but in every phase of ceremonial, military, and political life the 
moiety division appears more important than the clan division. 
The Chief moiety is associated with peace, the Warrior moiety 
with war. From among the Chiefs were chosen the highest public 
officials, while the Warriors took the initiative in setting out 
against the enemy. At the tribal ceremony the Chiefs cared for 
the medicine plants, while the Warriors presided over dances and 
games. As if to emphasize the distinctness of the moiety from the 
clan, there is a tendency — though not consistently carried out — 
for Chiefs to marry their daughters to other Chiefs rather than to 
Warriors.^^ In considering the social organization of the Yuchi, 
it would be monstrous one-sidedness to disregard the important 
dual division and emphasize only the exogamous clan unit. 

But the social relations of individuals may be definitely de- 
termined, even without a definite grouping under the same group 
name. Among the Hidatsa there were formerly seven exogamous 
clans, four of which were grouped in one moiety, and the re- 

^° Frank G. Speck, Ethnology of the Yuchi Indians, University of Pennsylvania, 
Anthropological Publications, University Museum, I (1909), asp. pp. 70-78. 


Kinship and Social Organization 

mainder in the complementary moiety. Membership in clan and 
moiety depended on one's mother's aflBliations. But a Hidatsa's 
social activities were only in part dependent on his membership in 
the two units mentioned; in very large measure they depended 
on his father's membership. The ceremonial life of an individual 
was determined by his father's: he was obliged to purchase cer- 
tain sacred objects from his own father, and with them member- 
ship in an esoteric ceremonial fraternity. Moreover, all the chil- 
dren of fathers belonging to the same clan were united in a group 
of "joking relatives" whose privilege it was not only to play prac- 
tical jokes on any of the members of the group, but also publicly 
to upbraid any member offending against tribal custom. Of the 
Hidatsa, also, we may therefore say that their social organiza- 
tion is very imperfectly described by an account of the exogamous 
clan system: at any particular period the group of "joking rela- 
tives" and of the patrilineal esoteric brotherhood loomed as 
equally important social factors with one's own clan. 

These few illustrations of different types of social units by no 
means exhaust the number found in North America. From among 
the most interesting ones I may add the castes of the Northwest 
coast, where tribes are divided into Chiefs, Noblemen, Com- 
moners, and Slaves; and loose associations of clans into phratries 
with apparently few or no distinctive traits, which occur among 
the Crow. Instead of finding North American society built upon 
a single basis, the clan or gens (for in Morgan's scheme, I re- 
peat, the phratry or moiety is only a segmented clan or gens), 
we have met the following varieties of social unit: 

1. The exogamous clan or gens (Crow, Fox). 

2. The clan or gens of indeterminate character as to exogamy, 
because it forms part of a larger exogamous unit (Iroquois). 

3. The exogamous moiety composed of several clans or gentes 
( Iroquois ) . 

4. The non-exogamous moiety composed of several clans or 
gentes (Hidatsa). 

5. The non-exogamous moiety organized without relation to 
clans or gentes (Yuchi, Fox). 

6. Phratries of indeterminate character (Crow, Kansas). ; ■ 

7. Castes (Northwest coast). 


Social Organization 

8. Ceremonial bodies in which membership is determined by 
descent (Hidatsa). 

To these> on the basis of former considerations, we must add 
the family in more or less our sense of the term, for we found that 
both clan or gentile systems exist and where they do not exist 
there are social relations due to ties of blood-relationship inde- 
pendently of such systems. 

To regard the exogamous one-sided kinship group as the sole 
basis of social organization in North America is thus an unjustifi- 
able piece of anthropological folklore. 

Finally, we must take up the question, whether the kinship 
group is everywhere fundamentally the same. To a certain 
extent this has already been answered. For if, on the one hand, 
clans are exogamous as such, like those of the Crow, while on the 
other hand they are only derivatively exogamous, as parts of other 
units that are exogamous in their own right, then, of course, what- 
ever fundamental unity exists, exists, from Morgan's point of 
view, between the exogamous divisions and not between the di- 
visions called clans. And as a definite unit must serve some pur- 
pose, the non-exogamous clan must have had some other char- 
acteristic that puts it in a different class from the exogamous 
clan. This being so, we must repudiate as dogmatic the assump- 
tion that wherever clan exogamy is found with other features the 
exogamous feature is the historically earliest trait with which the 
other traits afterward become associated. It is entirely possible 
that the course of development may in some cases have been as 
follows : There may have been a ceremonial group ^° at the be- 
ginning, with a rule of descent hke that followed in the Hidatsa 
esoteric groups, viz., the rule that children buy their own father's 
medicines and membership. The ceremonial group will thus come 
to consist of a number of patrilineal families. We need only tlie 
additional step that marriage shall be tabooed among fellow- 
members — a step that has been taken in ceremonial associations 
of a different type within the same area — to have a typical 
gens, with exogamy as the final instead of the fundamental fea- 
ture of organization. 

*" Or a group constituted for some other purpose. I adopt an example patterned 
as closely as possible on actual facts. 


Kinship and Social Organization 

In the light of these considerations we shall not base any theory 
as to the fundamental unity, either historically or psychologically, 
between two given clan systems on the existence of exogamy in 
both, but shall take into account all the geographical, historical, 
and psychological factors that would be considered in any other 
ethnological comparison. Let us compare, for example, the clan 
concept of the Iroquois with that of the Crow. The Iroquois clan 
bears an animal name; has a distinctive set of personal names; 
and was once derivatively exogamous. The Crow clan is not 
named for an animal, but bears a nickname, such as "Sore-lip," 
"Tied-in-a-Knot," "Bad War Honors"; it has no set of personal 
names; and it is primarily exogamous, for it does not form part 
of a larger exogamous unit. This parallel becomes really signifi- 
cant when we view both systems in their geographic setting. 
Clans and gentes bearing names directly or indirectly referring 
to animals are found very widely distributed among the tribes 
east of the Mississippi and among the Southern Siouan tribes. 
The idea of nicknaming clans, gentes, or local bands is also defi- 
nitely localized among the tribes of the Northwestern Plains — 
Assiniboine, Dakota, Blackfoot, Crow. Practically all the tribes 
with animal-named divisions have associated with these divisions 
sets of distinctive personal names; among the Northwestern 
Plains tribes children are named either in commemoration 
of some exploit of a distinguished tribesman or according to a 
supernatural revelation, a method precluding clan sets of personal 
names. Finally, there is fair evidence of exogamy being a sec- 
ondary clan feature among a number of Eastern Woodland and 
Southern Siouan tribes. The Iroquois, Winnebago, and Kansas are 
positively stated to have had exogamous moieties; for the Osage 
the same condition seems to have held formerly; ^^ for the Omaha 
there is at least some indication of pristine exogamy in the moiety. 
While the evidence on this point is far from convincing, a sus- 
picion remains that in the area under discussion exogamy may 
have been primarily associated with the dual division. But in the 
Northwestern Plains area there is no reason for assuming that 
exogamy was anything but a phenomenon characteristic of the 
clan or gens, for the simple reason that the moiety does not 

^^ For this information I am indebted to Miss Gerda Sebbelov. 


Social Organization 

occur among the Crow and Blackfoot, and the phratries of the 
Crow are historically a later development than the exogamous 
clan. The clans of the Iroquois and Crow are therefore not only 
radically dissimilar, but it appears that they represent two types 
of social unit distinctive of certain definite geographical areas. 
If kinship groups are not fundamentally alike even within the 
same continent, they will, a fortiori, not be fundamentally alike 
in different continents, as Dr. Goldenweiser has shown by a com- 
parison of Australian and Northwest American kinship groups. ^^ 
The theory that clans or gentes conform to a single basic concept 
thus breaks down utterly. 

Our critique of Morgan has thus established the following con- 

1. Kinship groups tracing descent unilaterally are not found 
universally among primitive tribes. 

2. It is not proved that the North American gentes developed 
out of clans. 

3. Restrictions of marriage are not primarily determined by 
unilateral kinship groups, for they exist, on the basis of blood-re- 
lationship, where no such groups exist, and coexist where such 
groups do exist; kinship groups being absent precisely among the 
tribes of lowest culture (in North America). 

4. The exogamous kinship group did not form the sole founda- 
tion of the social fabric among primitive tribes, where quite dif- 
ferent units, such as the moiety, caste, etc., occur, often coexist- 
ing with the clan or gens. 

5. The kinship group is a phenomenon of variable significance. 

Every destructive criticism of a view sanctioned by tradition 
leaves its adherents with a sense of loss. This feeling is of course 
an illusion, for there is no real loss when opinions are abandoned 
that are demonstrably false. But does modern ethnology, to use a 
hackneyed phrase, merely tear down without building up? To 
anyone acquainted even with the rudiments of psychology this 
question must seem very na'ive. Synthesis is the most fundamental 
characteristic of consciousness: all the elementary operations of 
the human mind, such as the association of ideas, are described by 

^^ Goldenweiser, op. cit., pp. 179-293. 


Kinship and Social Organization 

psychologists, such as HoflFding and Cornelius, as synthetic proc- 
esses. Under these circumstances the fear that modem critical 
science will compile a chaotic mass of data by failing to correlate 
the isolated facts it ascertains is absurd. The kind of synthesis 
that some people dream of seems to be the method of insisting 
that things are alike which one knows to be di£Ferent, and it 
is this kind of synthetic sleight-of-hand performance with which 
modern science will have nothing to do. Let us once more turn 
to the facts to see in what manner legitimate synthesis proceeds. 

We find that the Crow are divided into clans of a certain type 
— an isolated fact. When we find subsequently that the Hidatsa 
have exactly the same type of clan we spontaneously bring this 
fact under the same heading as the first — we synthetize, in other 
words, our two findings. This spontaneous activity, owing to the 
very nature of human consciousness, must be supplemented by a 
systematic attempt at all conceivable correlations of fact. How did 
the Crow-Hidatsa clan originate? Is it a trait of human psychology 
to evolve just such a clan in every part of the globe? Or is such 
a development due to a peculiarity of the Crow and Hidatsa? 
Or has it been borrowed from certain other tribes, and if so, from 
what tribes? The formulation of every such problem is the syn- 
thesis of certain facts that were at first isolated, unsynthetized. 
But the synthesis consists solely in the formulation itself. If we 
find that a suggested correlation does not hold as a matter of fact, 
our method of procedure has nevertheless been synthetic, quite 
independently of the result. When Morgan assumes without ques- 
tion that the Iroquois clan and moiety are at bottom alike, he is 
not more synthetic but less synthetic than we who question the 
self-evident character of his assumption and thus suggest a new 
coordination of the facts. 

The actual course of scientific progress is more enlightening 
on the subject of scientific method than the psychologically in- 
telligible, but logically often indefensible, desire for simplicity. 
We must indeed seek the simplest, most economical representa- 
tion of the facts, but the emphasis must be as strong on "facts" 
as on "economical" if we are to avoid producing merely a pleasing 
myth instead of a summary of reality. What does it mean when 
Faraday tells us that he crushed dozens of hypotheses in tlie 


Social Organization 

silence of his laboratory? It means that, thoroughly saturated 
with the facts of his science, Faraday attempted all manner of 
correlations of phenomena, many of which could never have 
suggested themselves to one unacquainted with the same range 
of facts, and that, by a process of selection, those correlations 
suggested in his thinking to which there corresponded a real 
correlation in nature remained as his permanent contribution to 

If anthropology is to be regarded as a science, it must conform 
to the logical methods of the exact sciences. Spontaneous synthesis 
of anthropological facts will be supplemented by systematic sug- 
gestion, verification, and elimination of all conceivable coordi- 
nations. In this work there will naturally be differences among 
anthropologists due to individual differences in knowledge of an- 
thropological fact, individual differences in degree of synthetic 
faculty and of critical judgment. But can anthropology aspire 
to the exactness of sciences like physics and chemistry? As re- 
gards logical method, at least, I firmly believe it can; if I did not, 
I should regard it as a harmless mode of amusement or as a 
branch of belles-lettres rather than as a branch of science. 

One or two illustrations must suffice. Among several of the 
Plains Indian tribes there are graded clubs, the members of any 
one of which are of the same age and buy their place in the so- 
ciety. We have here, then, two factors on either of which member- 
ship may primarily depend — purchase and age. The question is, 
which is the fundamental correlation, membership and purchase, 
or membership and age? Let us make an experiment in thought. 
Provided membership is a function of age, then if we vary the 
age beyond a certain limit membership must cease. Provided 
membership is a function of purchase, it will cease only when sold. 
Fortunately, our experiment in thought has been performed for 
us in reality. Among the Hidatsa and Mandan, individuals have 
preserved their membership in clubs regardless of advancing 
age, because they had had no opportunity to sell their member- 
ship; in this way some claim simultaneous aflBliation with more 
than one organization, which of course would be impossible if 
these clubs were primarily age societies. Hence, membership is 
basically a function of purchase. Many theories of physicists as to 


Kinship and Social Organization 

the constitution of matter rest, I fear, on a less solid basis of fact 
and logic. 

To take another case. Many primitive tribes use the same term 
in addressing relatives very differently related according to our 
notions of kinship. For example, the Crow call a father, a father's 
brother, the husband of a mother's sister or of a father's sister 
by the same term. They also use a single word in addressing a 
mother, a mother's sister, a father's sister, and the wife of a 
father's brother. From this we might infer that persons of the 
same generation and sex are addressed by the same kinship term. 
But while this conclusion is in large measure true, it does not 
state the whole truth. For a Crow will also address any father's 
clansman, no matter how young, as "father." Similarly, he will 
address any female of his father's clan, no matter how young, 
as his "mother." Now, as among the Crow descent is traced in 
the female line, any man, his sister, and her daughters and sons, 
her daughter's daughter, and all female descendants indefinitely 
must belong to the same clan. Hence, if I am a Crow, I ought 
to call my father's sister's daughter and all her female descendants 
"mother," and my father's sister's son "father," because all of them 
belong to my father's clan. This is actually the case, both among 
the Crow and the related Hidatsa. But how can we make sure that 
they are so called because of their clan? Obviously we must 
proceed again by keeping other possible causes constant and de- 
termining what happens when we eliminate the one cause under 
discussion. We must, therefore, find two relationships differing 
only as to clan affiliation. Now% while my father's sister's daughter 
and son belong to the same clan ( which is also my father's ) , the 
daughter of my father's sister's son and the daughter of my father's 
sister's daughter will belong to different clans, since the former 
must follow her mother's clan. Hence we have here two relatives 
identical in sex and generation, but differing in clan. We have 
found that a Crow will call his father's sister's daughter's daughter 
"mother." Does he call his father's sister's S07is daughter by the 
same term? He does not, calling her instead his "sister." In other 
words, when the clan factor is eliminated, a different factor be- 
comes potent — the generation factor. For it seems clear that I 
use the term "sister" because the father of my "sister" is my 


Social Organization 

"father" and because children of the same "father" are brothers 
and sisters. 

There are no bounds to the synthetic coordination of ethnolog- 
ical facts along the lines here indicated. That, however, ethno- 
logical facts should admit of such simple wholesale summing up 
as certain (not all) groups of physical facts is an unreasonable 
demand. If ethnological laws of development exist, their dis- 
covery will doubtless be a great achievement. But we must be on 
our guard against "fake" laws that do not result from a synthesis 
of the facts but from an artificial simplification by selection of 
those facts that fall in with the investigator's fancies. It would 
be a great simplification of much physical calculation if bodies 
fell with an acceleration of 10 meters per second; but they per- 
versely persist in falling with an acceleration of 9.81. So it would 
be charming if all tribes passed first through a loose, then through 
a clan, and finally through a gentile form of organization; un- 
fortunately there is no evidence that many of them do. It has 
been said in Newton's praise that he did not attempt to astonish 
himself by his clever ideas about nature, but sought to know what 
nature was really like. In looking over ethnological literature 
we are frequently tempted to ask whether the writer's object 
is not solely to amaze himself and others by his own cleverness. 
But ethnology is rapidly coming of age, and we are, learning to 
synthetize after the manner of Newton, in the expectation of rais- 
ing, not a structure of new anthropological folklore, but a new 
anthropological science. 



Exogamy and the Classijicatory 
Systems of Relationship 

Some connection between exogamy and primitive 
relationship terminologies has been recognized for a long time. 
Morgan noted that among the Iroquois all clan members were 
brothers and sisters as if children of the same mother.^ And though 
in his theoretical treatment of the subject he does not derive 
the classificatory system as a whole from the exogamous princi- 
ple, he does attribute the change from the older Malayan to the 
later and more common Turanian form of the system to punaluan 
marriage as a predecessor of the institution of exogamy and to 
exogamy itself." Tylor, to my knowledge, was the first to view 
exogamy and the classificatory system as but "two sides of one 
institution." ^ More recently both Frazer ^ and Rivers ^ discovered 

A paper read before the American Anthropological Association at Philadelphia. 
Printed in American Anthropologist, XVII (April-June, 1915), 223-239. 

^ L. H. Morgan, League of the Ho-de-no-sau-nee or Iroquois (New York, 1922), 
Book I, chap. 4. 

"Idem, Ancient Society (New York, 1877), Part III, chaps. 1-3. 

^ E. B. Tylor, "On a Method of Investigating the Development of Institutions, 
Apphed to Laws of Marriage and Descent," RAI, Journal, XVIII ( 1889), 245-269, 
esp. pp. 261 ff. 

* J. G. Frazer, Totemism and Exogamy (London, 1910), IV, 114. 

^ W. H. R. Rivers, "On the Origin of the Classificatory System of Relationships" 


Exogamy and Classificatory Relationship Systems 

the origin of the classificatory system in "a social structure which 
has the exogamous social group as its essential unit," both conceiv- 
ing this group as an exogamous moiety, which indeed had already 
figured prominently in Tylor's essay. 

It will be best to put this theory in somewhat more concrete 
form. Among the Iroquois, Morgan noted that a single term was 
applied to the maternal grandmother and her sisters; to the 
mother and her sisters; to the father and his brothers; and so 
forth. On the other hand, distinct terms were applied to the 
father's and to the mother's brother, to the father's and to the 
mother's sister.^ All these facts are readily formulated by de- 
riving the classification from the exogamous groups extant among 
the Iroquois: those relatives distinguished in our own nomen- 
clature and not distinguished in that of the Iroquois are mem- 
bers of the same exogamous division, while those not distinguished 
by us and separated by the Iroquois are necessarily members 
of different divisions. From this point of view the objection 
otherwise plausibly urged against denying the name "classifi- 
catory" to our own system since it, also, ranges certain relatives 
in classes becomes impossible. It is no longer a question, whether 
our terms "uncle," "aunt," or "cousin" are "classificatory" in a 
purely etymological sense of the term; nor whether the classifica- 
tory principle is quantitatively more important in certain primi- 
tive systems than in our own.^ The point at issue is the basis of 
the classification, and having regard to this there obviously exists 
a real difference between a system that classifies, say, cousins 
from both the father's and the mother's side under a common term 
and a system that rigorously divides relatives of the paternal and 
the maternal line on the ground of their different clan or gentile 
affiliations. Thus, the Tylor-Rivers theory, on the one hand, 
briefly summarizes and makes intelligible certain modes of classi- 
fication operative in many primitive systems that otherwise might 
seem purely capricious; and, moreover, it furnishes at last a 

in Anthropological Essays Presented to Edward Burnett Tylor (Oxford, 1907), 
pp. 309-323; idem. Kinship and Social Organisation (London, 1914), pp. 70 ff. 
The latter is an astonishingly stimulating contribution to tlae whole subject of kin- 
ship nomenclature. 

" Morgan, loc. cit. 

' Rivers, Kinship and Social Organisation, p. 2. 


Kinship and Social Organization 

logical basis for separating our civilized system from that of the 
primitive peoples concerned. 

I am profoundly impressed with the influence of the exogamous 
principle on primitive kinship nomenclature, but I feel strongly 
that the principle has not yet been formulated with adequate 
precision and due regard to coordinate principles of a different 
type. It seems to me that most writers on the subject suffer from 
the familiar disease of conceptual realism: the concept "classifica- 
tory system" is for them a sort of Platonic idea in the essence of 
which particular systems of this type somehow participate. 

Even Dr. Rivers cannot be freed from this charge. In practice 
he does not treat classificatory systems as fully determined by the 
clan or gentile factor, but smuggles in additional elements that 
go hand in hand with exogamy in moulding relationship nomen- 
clature. Most important among these is the principle, so strongly 
emphasized by Cunow, that members of the same generation 
are classed together and apart from other generations. Dr. Rivers 
regards this as so general a feature of "classificatory" systems that 
departures from the rule at once elicit from him special hy- 
potheses.^ Again, in a concrete illustration of his theory, he has it 
that in tribes possessing a classificatory system a person will apply 
a single term to all the members of his father's clan of the same 
generation as his father. I am certainly in favor of considering 
clan, generation, and other causes as jointly operative in the de- 
velopment of kinship nomenclature, but if this method is accepted 
an attempt must be made to indicate the interaction of these 
several principles. The fact is that the mode of interaction for the 
two factors that are here taken into account varies. In some 
cases, Dr. Rivers' statement, that members of the same clan (or 
gens) and generation are united, holds. But in other cases, for 
example among the Tewa, Crow, Hidatsa, and Tlingit, the exoga- 
mous principle predominates and overrides the generation cate- 
gory. Here, then, is an empirical problem, to be settled for every 
people and only obscured by the characterization of classificatory 
systems generally as "clan" systems; to wit, the problem how the 
exogamous group is coordinated with other principles of classifi- 

« Ibid., pp. 29-31. 


Exogamy and Classificatory Relationship Systems 

The fact that Dr. Rivers has not attempted to evaluate the sev- 
eral factors that together determine primitive relationship ter- 
minology has led him into the curious position of underestimating 
in practice the very factor that occupies the dominant position in 
his theory. Again and again he invokes special social usages to ac- 
count for "relatively small variations of the classificatory system" 
that are at once explained by the prepotency of the exogamous 
principle. For example, he cites the East Indian term bahu, which 
is applied to the son's wife, the wife, and the mother; and in ex- 
planation of this classification he assumes a one-time form of 
polyandry in which a man and his son had a wife in common.^ 
This assumption is de trap because with a dual organization and 
paternal descent, I and my son belong to my father's moiety, 
while my mother, my wife, and my son's wife must belong to 
the complementary moiety; hence, bahu may simply connote 
females of that exogamous group. Again, Dr. Rivers cites the 
Pawnee use of one term for the wife and the wife of the mother's 
brother, explaining tliis by a special form of marriage. ^*^ But, given 
a dual organization with maternal descent, I and my mother's 
brother are members of the same moiety, while my wife and his 
wife are fellow-members of the complementary moiety. Finally, 
the confusion of generations in the Banks Islands ^^ requires no 
special hypothesis. With maternal descent, my father's sister's 
son is classed with my father because, as among the Tewa of 
Hano, he is my father's clansman. My mother's brother's children 
are classed with my children because my mother's brother, being 
my clansman, is my brother; and because two brothers regard 
each other's children as their own. Thus, in the first confusion of 
generations the clan principle alone has been operative; in 
the second case, the clan principle has established a relationship 
from which a really non-existent distinction of generation is the 
logical derivative. 

Thus, on the one hand, the exogamous theory does not sufiice 
to explain the "classificatory" systems in their totality; on the 
other hand, it eliminates certain auxiliary hypotheses considered 

" Ibid., p. 90. 

'" Ibid., pp. 53-54. 

" Ibid., pp. 28 ff. 


Kinship and Social Organization 

necessary by the most eminent of its advocates. Obviously, there 
is something u^rong with the formulation of the theory. 

The solution of the difficulty lies implicitly in the original and 
invaluable portion of Professor Kroeber's essay on the subject ^^ 
— the very part ignored or misunderstood by his critics — where 
he lists the categories in which the North American Indians have 
ranged relatives, and by the quantitative importance of each of 
which a given system may be defined. Among these categories is 
that of distinguishing lineal and collateral relatives, — the father 
from the father's brother, the mother from the mother's sister, 
the brother from the cousin. When we turn to Morgan's earliest 
description of what he afterwards took for the starting-point of 
his definition of the classificatory type of system,^^ we find that 
what impressed him above all was the abeyance of the rule that 
collateral shall be distinguished from lineal relatives. This, then, 
forms the core of Morgan's concept, however obscured by ad- 
herent features that are logically quite unrelated. And from this 
point of view the Tylor-Rivers theory assumes a different aspect. 
Exogamy cannot explain why generations are so generally dis- 
tinguished; it cannot explain the frequent differences between 
elder and younger Geschwister, or the frequent distinction be- 
tween vocative and non- vocative forms; it cannot explain a hun- 
dred and one features of classificatory systems so-called. But it 
does explain why lineal and collateral lines of kinship are merged 
in the particular way characteristic of the Iroquois, Ojibwa, and 
many other primitive systems conforming to Morgan's Turanian 
type. Thus purged, the theory must now be subjected to empirical 

It might appear at first sight that such an empirical verifica- 
tion has already been given by Dr. Rivers with regard to Oceania, 
though this, of course, would not render it unnecessary to collect 
corroborative evidence from other regions. However, Dr. Rivers 
has in reality made a different point. In Oceania he is not dealing 
with classificatory and non-classificatory systems, but merely with 
the two forms of the classificatory system, — the Hawaiian and the 

" A. L. Kjoeber, "Classificatory Systems of Relationship," RAI, Journal, XXXIX 
(1909), 77 ff. 
^ Morgan, op. cit., Book I, chap. 4. 


Exogamy and Classificatory Relationship Systems 

Turanian. In both forms lineal and collateral relationship are 
merged, but in the Hawaiian nomenclature the terms are even 
more inclusive, no distinction being drawn between relatives of 
the maternal and the paternal side. Setting out with his theory 
and these two forms of the classificatory system before him, Dr. 
Rivers undertakes to show how the Hawaiian form could have 
developed from the Turanian form which alone follows logically 
from the exogamous principle. He finds that where the Hawaiian 
form is most clearly developed, traces of exogamy are lacking, 
while the highest development of exogamy is accompanied by a 
Turanian form of kinship system. An indefinite number of inter- 
mediate social organizations are accompanied, we are told, by 
intermediate kinship terminologies. The, general interpretation of 
the phenomena offered is that a progressive change has occurred 
from the Turanian to the Hawaiian form, going hand in hand with 
the substitution of non-exogamous marriage regulations for regu- 
lation by exogamous divisions.^"* 

This is obviously not testing the theory that the classificatory 
system is a function of exogamy, but merely interpreting by a 
special historical hypothesis the occurrence of an aberrant type 
of classificatory system on the supposition that the theory is al- 
ready established. Granting that the hypothesis correctly repre- 
sents the course of development in Oceania, we cannot assume 
that exogamy everywhere represents an older condition, and in- 
deed in North America the evidence points in the opposite direc- 
tion.-^^ Without assuming the priority of either the exogamous or 
the loose social organization, we can test the Tylor-Rivers theory 
by grouping together exogamous tribes, on the one hand, and non- 
exogamous tribes, on the other hand, and comparing the corre- 
sponding kinship terminologies. North America, where the geo- 
graphical distribution of types of organization is fairly well de- 
termined, offers a favorable field for such an inquiry. 

In the first place, there can be little doubt that the custom 
of identifying in nomenclature lineal and collateral relatives is 
very largely coextensive with the exogamous practice. It is found 

** Rivers, Kinship and Social Organisation, pp. 65 ff. 

^^ See R. H. Lowie, "Social Organization," American Journal of Sociology, XX 
( 1914 ) , 68-97, reprinted as no. 1 in this volume. 


Kinship and Social Organization 

in at least tliree of the four main exogamous areas of the continent, 
— east of the Mississippi, among the southern Siouan and north- 
western Plains tribes, and on the Northwest coast. For the South- 
west satisfactory data seem to be lacking except for the Tewa,^^ 
where conditions are markedly anomalous. Both among the pat- 
rilineally organized Tewa of New Mexico and the matrilineal 
Tewa of Hano, Arizona, there are distinct terms for father's 
brother and father, although the term for "father" is "applied 
loosely to father, elder brother, father's brother, or other relatives 
older than self" in New Mexico,^^ and at Hano to all father's 
clansmen, including of course his own brothers. ^^ For the present 
purpose it is important to note that at Hano it is the distinct 
term for "father's brother" that seems to be the older mode of 
designation, now rendered obsolescent by the term for "father." ^^ 
Similarly, in both branches of the Tewa, the mother's sister is 
carefully distinguished from the mother; in Hano, we are told em- 
phatically, a mother's sister is never addressed as mother; and 
conversely we find that a woman does not address her sister's 
children like her own children but by a reciprocal term with 
diminutive suffix.-*^ In New Mexico there is a further invasion of 
the exogamous principle inasmuch as no distinction is drawn be- 
tween paternal and maternal uncles and aunts respectively.^^ 
For the Tewa, then, the hypothetical correlation does not hold. 
In the New Mexican division of the people the grouping of rela- 
tives has been affected only to a very slight degree by the gentile 
organization. At Hano the effect of the corresponding clan or- 
ganization has been greater, for among other extensions, the word 
for "child" is applied by a male to any of his clansmen's children 
and a single term embraces all the speaker's clansmen other than 
his own brothers. Nevertheless, even in this pueblo the diver- 
gence from the type in the designations for father, mother, and 

"J. P. Harrington, "Tewa Relationship Terms," American Anthropologist, XIV 
(1912), 472-498; Barbara Freire-Marreco, "Tewa Kinship Terms from the Pueblo 
of Hano, Arizona," ibid., XVI (1914), 269-287. 

" Harrington, op. cit., p. 479. 

" Freire-Marreco, op. cit., p. 277. 

" Ibid., p. 278. 

"^ Ibid., p. 276; Harrington, op. cit., p. 488. 

^ Harrington, op. cit., pp. 487, 488. 


Exogamy and Classificatory Relationship Systems 

child is so great that we cannot, without doing violence to the 
facts, describe the kinship system as a "clan" system. In short, 
here is a striking instance of exogamy without an exogamous 
alignment of kindred.^" 

A thoroughgoing explanation of the Tewa anomaly will be pos- 
sible only when other southwestern systems shall have become 
known. For the present a few hints must suffice. So far as the 
Hano people are concerned, it will be well to remember that their 
present rule of descent, as well as parts of their kinship ter- 
minology, may be due to the influence of the surrounding Hopi 
and their isolation from the other Tewa pueblos. Not the in- 
habitants of the single village of the Tewa enclave in Hopi ter- 
ritory, as Miss Freire-Marreco seems to suppose, but their New 
Mexican congeners may use a kinship system approaching the 
ancient Tewan type. The problem thus narrows down to that of 
explaining why the patrilineal Tewa use a nomenclature that 
does not reflect their gentile grouping. Two alternative solutions 
occur to me. Either the Tewa adopted their present social or- 
ganization at so recent a period that the innovation has not yet 
affected their mode of designating relatives. Or they have aban- 
doned an older kinship system and borrowed a new one from 
some non-exogamous tribe. A third possibility, however distaste- 
ful to some minds, must be reckoned with. Though the influence 
of clan or gentile organization on kinship terminology seems to be 
a very general phenomenon, it cannot be accepted as a law of 
nature. It remains conceivable that a tribe should possess an 
exogamous social organization that finds little or no expression 
in the linguistic designation of kindred, just as it is now an es- 
tablished fact that the linguistic grouping of different relatives 
under the same category does not blind the users of such a ter- 
minology to the differences in the relationships. 

However this may be, the single exception that has been noted 
cannot invalidate the empirical rule tliat there is a considerable 
correlation between exogamy and the merging of lineal and col- 
lateral relatives. Our next question is, what happens to kinship 

^ This is confirmed by Morgan's statement that in Tesuque, the southernmost 
Tewa pueblo, all cousins alike were addressed as brothers and sisters. L. H. Morgan, 
Systems of Consanguinity and Affinity (Washington, 1871), p. 263. 


Kinship and Social Organization 

nomenclature among non-exogamous tribes? The most important 
North American tribes without a clan or gentile organization 
are the Eskimo, the Mackenzie River, Plateau, and California 
Indians. Let us rapidly survey representative kinship terminolo- 
gies of these tribes. 

The Eskimo system differs so radically from the characteristic 
"classificatory" form that even Morgan claimed at best only a 
remote relationship between the two. The father's brother is dis- 
tinguished from the father, the mother's sister from the mother, 
the children of my father's sister and of my mother's sister are 
designated by a common term, and so forth. ^^ For the Mackenzie 
River district Morgan's informants seem to establish the presence 
of fairly well defined exogamous kinship features among these 
non-exogamous peoples; however, in contravention of such a 
system cousins are uniformly addressed as brothers and sisters.^^ 
Moreover, the exogamous features may be the result of borrow- 
ing from two distinct sources, — the Algonkian tribes to the south, 
and the Pacific coast population to the west. The striking coin- 
cidence of certain Northern Athapascan with Algonkian traits is 
noted by Morgan himself. Passing to the Salish tribes of the in- 
terior of British Columbia, we find a marked departure from the 
exogamous type of nomenclature. The Coast Salish draw no dis- 
tinction between cousins on the father's and the mother's side; 
class together paternal and maternal uncles and aunts, distin- 
guishing them from the parents; and distinguish children from all 
nephews and nieces. The Bella Coola likewise distinguish uncles 
and aunts from parents, and class together those of the paternal 
and those of the maternal line. Among the Shuswap there seem 
at first sight to be some "classificatory" features inasmuch as 
nephews are classed with sons, and nieces with daughters; but 
they are classificatory only in an etymological, not in Dr. Rivers' 
sense of the word, since the brother's and the sister's children are 
included in the same category. Boas points out that, while the 
Shuswap distinguish the parents from their brothers and sisters. 

^ Morgan, Systems, pp. 276-277. 

'^ Ibid., pp. 234-340. In Chipewyan, according to Le Goff (Gramnmire de la 
langue montagnaise, p. 330), a distinction as to father's and mother's side is made 
when the speaker addresses a cousin of the opposite sex. 


Exogamy and Classificatory Relationship Systems 

the term used by boys for uncles coincides with the stem for 
"father" in other dialects, while that used by girls for aunts ap- 
proximates the stem for "mother" in other branches of the Salish 
stock. This fact is of the utmost significance for a study of kinship 
terminology from a psychologico-linguistic point of view, but has 
no bearing on the present issue since, even in these instances, the 
uncles and aunts are not distinguished as to paternal or maternal 
side. Among the Okanagan this difference is indeed made, but the 
terms for parents remain distinct. In Kalispel Salish, likewise, the 
mother's sister is not confounded with the mother in nomencla- 
ture, nor the father's brother with the father; the terms applied 
by a female to her nephews and nieces are obviously related to 
those for "son" and "daughter," but again the brother's and sis- 
ter's children are not distinguished. In short, the non-exogamous 
Salish tribes have a non-exogamous kinship system. ^^ For Cali- 
fornia published data are meager, but Kroeber's statement that 
the systems of that area display a remarkable differentiation be- 
tween the lineal and collateral lines supports the assumption 
of their non-exogamous character.^^ Finally, we may consider the 
Shoshonean family for which partial lists by Sapir are available 
for the Kaibab Paiute and Uintah Ute,^^ as well as unpublished 
data collected by the present writer among the Wind River Sho- 
shone, White River ( ? ) Ute, Southern Ute, and Northern Paiute, 
not to forget Morgan's imperfect presentation of the Uncom- 
pahgre system. With the single exception of the Wind River Sho- 
shone, the kinship nomenclature of the entire stock is markedly 
non-exogamous: parents are distinguished from uncles and aunts, 
children from all nephews and nieces. Among the Kaibab mater- 
nal and paternal uncles or aunts are not distinguished, and even 
among the Wind River Shoshone all cousins are designated by 
a single term. On the other hand, all the Shoshonean systems 
are characterized by a feature shared with the Tewa, — the fre- 
quent use of reciprocal instead of correlative terms for the mem- 

^ Franz Boas, "Terms of Relationship of the SaUsh Languages," Report of the 
Sixtieth Meeting of the British Association for the Advancement of Science, 1890 
(London, 1891), pp. 688-692. 

^ Kroeber, loc. cit. 

^ Edward Sapir, "A Note on Reciprocal Terms of Relationship in America," 
American Anthropologist, XV (1913), 134 ff. 


Kinship and Social Organization 

bers of a related pair. We may well pause for a moment to con- 
sider the influence of historical accident on the course of theoreti- 
cal speculation. It has been said by an eminent scientist that if 
physicists had first studied thermal rather than mechanical phe- 
nomena, heat would not have been described as a mode of motion 
but motion as a mode of heat. Had Morgan begun his researches 
in the Plateau area, we might have heard less of classificatory 
and more of reciprocal systems of relationship. 

Summing up the result of our sketchy survey, we may say 
that the Tylor-Rivers theory derives strong corroboration from 
North American data. Despite some conflicting evidence exoga- 
mous kinship systems coincide so largely with an exogamous so- 
cial organization and are so commonly lacking where exogamy 
does not obtain that a functional relation between the two must 
be regarded as more than probable. 

We are thus emboldened to pursue our inquiry somewhat more 
rigorously. Having compared exogamous tribes as a whole with 
non-exogamous tribes as a whole, we may profitably undertake 
a more intensive comparison of narrower scope. Since a multi- 
plicity of operative causes must be recognized, it becomes neces- 
sary to minimize all other difi^erences save in point of exogamy for 
the purpose of studying the effect of that factor by itself. This may 
be done by grouping tribes according to various principles of 
classification. Within the Algonkian stock, for example, the Cree 
are reported to lack an exogamous organization. How, then, does 
the Cree kinship system compare with that of the linguistically 
and culturally most closely related Algonkian tribes possessing the 
gentile organization, such as the Ojibwa? From Morgan's data it 
appears that the kinship terminologies of these tribes agree very 
closely, both indicating the influence of the exogamous factor.^^ 
Does, then, the Tylor-Rivers theory break down? Not at all. It 
must simply be taken in connection with certain concrete facts. 
That kinship nomenclature may persist after the conditions in 
which it originated have disappeared, is a principle never urged 
more emphatically than by Morgan himself. The Cree termi- 
nology may therefore well be a survival from a former gentile 
organization, which is in this instance rather probable from its 

^ Morgan, Systems, pp. 204-208. 


Exogamy and Classificatory Relationship Systems 

prevalence in a number of closely allied Algonkian peoples. Again, 
within the same stock, the Arapaho and the Gros Ventre, sharing 
the essential traits of Plains culture and speaking mutually in- 
telligible dialects, differ in point of organization; the Arapaho 
having a non-exogamous and the Gros Ventre a gentile system. 
If the Arapaho and Gros Ventre kinship systems differed ac- 
cordingly, the history of the difference might be sought in the 
influence of the Blackfoot with whom the Gros Ventre have been 
in intimate contact and from whom they have demonstrably bor- 
rowed a number of cultural traits. The Arapaho and Gros Ventre 
systems, however, are identical, and, except for the designation 
of cousins, conform to the exogamous type.^^ Shall we be inclined, 
as in the previous instance, to assume a former condition of 
Arapaho exogamy? 

This raises an important problem. Tylor was so strongly im- 
pressed with the correlation of "classificatory" systems and ex- 
ogamy that he felt warranted in inferring exogamy from the 
presence of such a system. Nowadays, we shall hardly go quite so 
far. Granting that a correlation is established, it would still have 
to be proved that it is a hundred per cent correlation, that not 
only do exogamous peoples possess a corresponding kinship sys- 
tem but that no other cause could have produced such a system. 
The geographical distribution of certain relationship categories 
demonstrates that these categories, like other cultural traits, may 
be diffused by borrowing. In the paper already quoted Dr. Sapir 
points out that in the southwestern United States the use of 
reciprocal terms with diminutive suffixes to designate the junior 
relative is strangely similar among the Tewa and the Shoshonean 
tribes. The distribution of reciprocal systems is so definitely lo- 
calized in North America, without being confined to members of 
a single linguistic family, as to become intelligible only on the 
hypothesis of borrowing. This being so, the confusion in language 
of lineal and collateral relatives may likewise have spread through 
historical connection, as has already been suggested for the Atha- 
pascan of the Mackenzie River region. We shall therefore cer- 
tainly be on the alert for evidence of former exogamy where the 
kinship system is exogamous, but we shall not accept the system as 

'"'A. L. Kroeber, The Arapaho, AMNH Bull., XVIII (1902-1907), 9 f., 150. 



Kinship and Social Organization 

proof of exogamy unless, as in the Cree instance, there are spe- 
cific conditions to corroborate the conclusion. In the Arapaho 
case the conditions are not, in my opinion, of such a character. 
Here we have not a group of closely related tribes all of whom, 
with a single exception, possess an exogamous organization. The 
Arapaho and the Gros Ventre stand alone within the Algonkian 
family, and as already stated the Gros Ventre may have borrowed 
the gentile organization from the Blackfoot. Under the circum- 
stances I do not pretend to give a solution of the problem but 
content myself with enumerating various possibilities. It may be, 
as Tylor would argue, that the Arapaho-Gros Ventre originally 
had an exogamous organization still preserved by the Gros Ventre, 
of which the Arapaho kinship system is a survival. Or, the as yet 
undivided parent tribe lacked exogamy, but borrowed an ex- 
ogamous terminology from some neighboring people. Or, the 
parent tribe had neither exogamous divisions nor an exogamous 
nomenclature, but the Gros Ventre adopted them from the Black- 
foot; and the Arapaho, in recent times, borrowed the Gros Ventre 
terminology. Or, the common terminology developed quite in- 
dependently of exogamy, — to my mind the least acceptable hy- 

The two illustrations hitherto given of intensive comparison 
have been made on the basis, primarily, of linguistic affiliation. 
Important as such a classification must be when we are dealing, 
after all, with elements of speech, it is not always possible. This 
applies, for example, to California, where the degree of linguistic 
differentiation necessitates a diflFerent mode of grouping. Here 
we might ask, for instance, how the exogamous Miwok differ as 
regards kinship from their non-exogamous neighbors, such as the 
Maidu, Washo, or Yokuts. Similarly, in the Plains area, the sys- 
tem of the non-exogamous Kiowa would be of great interest for 
our present purpose. Again, the Pawnee seem to have an exoga- 
mous kinship system without exogamy. This may be a survival 
from one-time exogamy; but it may also be the result of borrow- 
ing, and a detailed comparison of the Pawnee system with that 
of all the tribes with which the Pawnee have come in contact 
may determine its source of origin. 

Only tlirough such intensive studies of detail shall we obtain 


Exogamy and Classificatory Relationship Systems 

an insight into the workings of the exogamous principle in its 
effect on kinship terminology and in its relations to other prin- 
ciples that may check or nullify its influence. 

As has already been suggested, exogamy may do more than 
produce the fusion of lineal and collateral lines; as among the 
Hano Tewa, Crow, and other tribes, it may override the genera- 
tion principle. This possibility is admitted by Dr. Rivers, who is, 
however, strongly inclined to explain the disregard of generations 
not by exogamy but by the practice of special forms of marriage. 
The special case of the father's sister's son being classed with the 
father is intelligible, we are told, if a man marries the wife or 
widow of his mother's brother, for thus he comes to occupy his 
maternal uncle's social status, and his uncle's children therefore 
regard him as their father.^*^ This assumption has already been 
criticised from the point of view of logical method: on the prin- 
ciple that hypothetical causes shall not be multiplied unnecessarily 
I have argued that no special hypothesis should be advanced for 
minor variations of the classificatory system if the theory pur- 
porting to explain that system as a whole suffices to explain the 
variations. This argument does not of course refute the existence 
of special causes. How, then, can we be sure that it is the exoga- 
mous factor and not some such social usage as that suggested 
by Dr. Rivers that determines the classification under discus- 
sion? In the first place, it should be noted that the neglect of 
generations among the Tewa, Hidatsa, and Crow is not limited to 
the person of the father's sister's child, but that a single term is 
applied to the father's sister and all her female descendants, im- 
mediate and through females, ad infinitum. Shall we construct 
successive hypotheses as to forms of marriage by which a man 
would become the son of his father's sister's daughter's daughter 
and her successive female descendants when all the facts are 
summed up by the plain statement that there is one word for a 
father's clanswoman? That the clan is indeed the determining 
factor, is indicated by the effect of eliminating it. Among the 
Crow, as soon as we pass out of the clan by taking the daughter 
of the father's sister's son rather than of the father's sister's daugh- 
ter, the generation factor at once enters: my father's sister's son 

^^ Rivers, Kinship and Social Organisation, pp. 54, 28-31. 


Kinship and Social Organization 

is indeed my father since he belongs to my father's clan, but his 
daughter belongs to her mother's clan, hence is related to me only 
genealogically as my "father's" daughter, hence is my sister, not 
my mother or aunt.^^ 

The conclusion is corroborated by other evidence. If the rule 
of descent is changed from maternal to paternal, kinship nomen- 
clature should not be affected provided it has been shaped by 
forms of marriage, which are substantially uniform throughout 
the Plains area. If, on the other hand, kinship nomenclature has 
been shaped by exogamy the change from matrilineal to patri- 
lineal descent ^^ must produce a change in the terminology since 
my father's sister's daughter no longer belongs to my father's 
exogamous division. Within the Siouan family, to which the mat- 
rilineal Crow and Hidatsa belong, there are also patrilineal tribes, 
of which the Omaha are the best known. Among the Omaha my 
father's sister's daughters are classed not with my father's sister 
but with my sister's children, while her husband is my brother- 
in-law. These facts seem to indicate the influence of the gentile 
factor, for my father's sister belongs to my gens and is there- 
fore my gentile sister, although it must be noted that the Omaha 
distinguish the father's sister from the own sister. The Crow and 
Hidatsa with greater consistency class not only the mother's 
brother's children with the brother's children — which corresponds 
exactly to the Omaha usage, having regard to the given differences 
in descent — but also class the mother's brother with the elder 
brother. Thus, the change from the clan to the gens has eliminated 
the classification of the father's sister's children with the first as- 
cending generation, and their classification with the first descend- 
ing generation among the Omaha becomes intelligible through the 
gentile principle. But this principle may be proved to have been 
operative in a quite unexceptionable manner. The Omaha class 
the mother's brother's son and all his male descendants, immediate 
or through males, with the mother's brother. And again as soon as 
we pass out of the exogamous group, the terminology varies: 

^^ The Crow address the father's sister as "mother," but refer to her non-\ocatively 
as "aunt." 

°" This phrase is not to be taken as expressing the chronological order of events 
but simply the conditions of our Gedankcnexpcrimcnt since we are passing from 
the consideration of a matrilineal to that of a patrilineal tribe. 


Exogamy and Classificatory Relationship Systems 

my mother's brother's son's son is my mother's brother, but my 
mother's brother's daughter's son is my brother since the bond be- 
tween him and me is no longer gentile but genealogical, through 
his mother, who is my "mother." ^^ According to Morgan these 
characteristics obtain for all the southern Siouan tribes. ^^ In other 
words, where the matrilineal Hidatsa and Crow Indians class in 
one category the father's sister and her female descendants, im- 
mediate and through females, the patrilineal Omaha, Oto, Kansa, 
and other southern Siouan tribes unite the mother's brother and 
his male descendants, immediate and through males, ad infinitum. 
Better proof could hardly be demanded for the theory that the 
disregard of generations is the result of the exogamous prin- 

Exogamy thus furnishes a su£Bcient explanation of the invasion 
of the generation principle as encountered in Melanesia and vari- 
ous North American tribes. The chief value of the theory that kin- 
ship classification has followed exogamous groupings lies, how- 
ever, in another direction. It explains the remarkable resemblance 
between the terminologies of widely separated and quite distinct 
peoples without recourse to hypothetical historical connections. 
If we abandon Morgan's theory that the development of the family 
has been unilinear, with the main stages impressing their stamp 
on kinship nomenclature, how can we account for the far-reaching 
similarity between, say, the system of the Seneca of North America 
and that of the South Indian Tamil? So widespread a custom as 
exogamy is admirably fitted to explain the distribution of the 
lineal-collateral category, and it seems eminently worth while 
to extend the verification of the Tylor-Rivers theory both ex- 
tensively and intensively. Such a study will be far from exhaust- 
ing the subject of kinship nomenclature. The merging of lineal 
and collateral relationships constitutes but one of a number 
of categories, the geographical distribution of each of which must 

^ J. O. Dorsey, "Omaha Sociology," BAE, 3d Ann. Reft. ( 1884), p. 254; Morgan, 
Systems, pp. 335-336. 

^ Morgan, Systems, pp. 178-179. 

^ In comparing the Choctaw with the Omaha system of kinship Dr. Kohler has 
called attention to the influence of the rule of descent on nomenclature. See Josef 
Kohler, "Zur Urgeschichte der Ehe," Zeitschrift fiir vergleichende Rechtswissen- 
schaft, XII (1897), 187-353, esp. p. 303. 


Kinship and Social Organization 

be definitely ascertained. Moreover, we are sadly in need of the 
intensive investigation of particular systems, giving all the con- 
notations of every term, and indicating by comparison with closely 
related systems how and why kinship nomenclature changes. A 
comparative study of all the Siouan, or all the Athapascan, or all 
the Southwestern systems would be of the greatest value in this 
respect. However, the connection between exogamy and the "clas- 
sificatory" system often hinted at but never systematically ex- 
amined before Dr. Rivers' investigations in Oceania constitutes 
even by itself a problem of great significance and its partial solu- 
tion cannot help but to react on a study of other phases of the 
whole question of kinship terminology. 


Historical and Sociological 
Interpretations of 
Kinship Terminologies 

Students of kinship terminology have been interested 
almost exclusively in the sociological inferences that may be de- 
rived from systems of relationship. They have generally failed to 
note that Morgan himself drew not merely sociological, but also 
startling historical conclusions from the observed phenomena. In- 
deed, in this regard Morgan may fairly be said to out-Graebner 
Graebner. He not only rejects the hypothesis that similarities in 
relationship nomenclature can be explained by independent de- 
velopment, but also summarily dismisses the suggestion of dif- 
fusion by borrowing: nothing will do but racial aflBliation. 

In other words the Turanian and Ganowanian famiUes drew their 
common system of consanguinity and affinity from the same parent 
nation or stock from whom both were derived, etc. . . . When the 
discoverers of the New World bestowed upon its inhabitants the 

Holmes Anniversary Volume: Anthropological Essays Presented to William 
Henry Holmes . . . December 1, 1916, by His Friends and Colaborers (Wash- 
ington, 1916), pp. 293-300. 


Kinship and Social Organization 

name of Indians, under the impression that they had reached the 
Indies, they little suspected that children of the same original family, 
although upon a diflFerent continent, stood before them. By a singular 
coincidence error was truth. ^ 

This extravagant view was quite correctly criticized by Lub- 
bock when he pointed out that the Two-Mountain Iroquois can 
hardly be recognized as more closely akin to remote Oceanian 
tribes than to their fellow-Iroquois. But while this criticism (with 
an indefinite number of similar instances that lie at hand ) elimi- 
nates the use of kinship terminologies for ascertaining racial affin- 
ity, it does not dispose of them as evidence of cultural connection. 
Indeed, Morgan himself repeatedly cites resemblances that be- 
come intelligible only from this point of view; yet with that char- 
acteristic lack of the logical sense that detracts so largely from his 
otherwise superb pioneer achievements he fails to see the bearing 
of his data. Professor Kroeber — the only writer since Morgan who 
has departed fundamentally from the sociological point of view 
— assumes an historical position when he points out that dif- 
ferences in terminology are regional; but so far as I can see, he 
does not stress the obvious conclusion that the similarities within 
a given region are due to historical connection.^ 

In the following pages I will cite a number of striking similarities 
that are explicable as the result of historical connection and will 
discuss the relation of these facts to a sociological interpretation. 

While the differentiation of elder and younger brothers and sis- 
ters is of very common occurrence, a tripartite classification of 
Geschioister is not found, so far as I know, except among the 
Eskimo. The Alaskan Eskimo, according to notes supplied by Dr. 
E. W. Hawkes, have distinct terms for elder brother, younger 
brother, and youngest brother. Corresponding to this, we find 
among the Chukchee three distinct terms for eldest brother, mid- 
dle brother, and youngest brother; and a similar nomenclature 
among the Koryak.^ The peculiarly restricted distribution of the 

^ L. H. Morgan, Systems of Consanguinity and Affinitt/ (Washington, 1871), 
p. 508. 

^ A. L. Kroeber, "Classificatory Systems of Relationship," RAI, Journal, XXXIX 
(1909), 81. 

^'M. A. Czaplicka, Aboriginal Siberia (Oxford, 1914), pp. 30, 35. 


Interpretations of Kinship Terminologies 

phenomenon is such as at once to suggest diflFusion. Here, how- 
ever, an alternative explanation may be given. Though dijffusion 
be the ultimate cause, the immediate antecedent of the similarity 
may be similar social conditions. In other words, it is certain 
social peculiarities relating to the status of eldest and youngest 
brothers that may have been borrowed and the kinship terms may 
have developed independently from these borrowed usages. 

A like perplexity confronts us when we consider the systems 
found east of the Mississippi and some of the adjoining Plains 
territory to the west. It is true that a characteristic trait of these 
systems — the merging of collateral and lineal kin — occurs in a 
continuous, though vast, area so that its distribution would, ac- 
cording to accepted criteria, be explained by dissemination from 
a single source. However, it is also true that this area is practically 
coextensive with the Eastern area of clan exogamy. The hypoth- 
esis is, therefore, a priori tenable that clan exogamy was the 
feature diffused, from which a corresponding kinship nomencla- 
ture developed independently in a number of cases. We must 
eliminate such possibilities if we are to establish the historical 
significance of kinship terms themselves. 

Among the numerous tribes in the eastern and central United 
States which designate collateral kin by the terms used for lineal 
relatives there are nevertheless certain far-reaching differences 
connected with the tendency to recognize or to ignore diflFerences 
of generation. Mr. Leslie Spier's as yet unpublished researches 
in this field indicate that this tendency appears primarily in the 
designation of cross-cousins. This phenomenon had not escaped 
the attention of Morgan.* As he points out, the Seneca, Ojibwa, 
and Dakota designate a mother's brother's and a father's sister's 
child as "cousin"; the Southern Siouans (and Winnebago) call 
the mother's brother's son "uncle" and the father's sister's son 
"nephew"; the Crow and Hidatsa class the former with the son 
and the latter with the father: and at least part of the Crow- 
Hidatsa scheme occurs among the Choctaw and related tribes.'' 
I have elsewhere shown that the striking diflFerence between 
the Southern Siouans and their northern congeners, the Crow 

Mor2;an, op. cit., p. 189. 

For the sake of simplicity I cite only some of the relevant data. 


Kinship and Social Organization 

' and Hidatsa, is a function of the difference in rules of descent.^ 
Owing to the very close linguistic affiliation of the Crow and 
Hidatsa, most ethnologists will agree that their kinship systems 
( which coincide in many points besides those cited ) were neither 
borrowed from each other in recent times nor arose independently, 
but are survivals from the system of the parent tribe, which 
pristine system, then, expressed the maternal (clan) organiza- 
tion. For like reasons, a corresponding conclusion will be deemed 
permissible for the Omaha and their immediate relatives, whose 
pristine system, as already noted by Kohler and Cunow, re- 
flected their paternal (gentile) organization. 

But as the example of the matrilineal Seneca and the patri- 
lineal Ojibwa indicates, the rule of descent need not be reflected 

- in the kinship nomenclature; for here we have tribes of dif- 
ferent rules of descent with the same mode of designating cross- 

• cousins, who are not placed in a generation above or below that 
of the speaker. I will not now attack the more general problem 

\ why certain tribes emphasize rules of descent while others do not. 
I prefer to render the problem more specific, and therefore more 
amenable to solution by confining attention to a single linguistic 
stock and to a single branch of that stock, among whose members 
there is a cultural bond as well. Of the Central Algonquians, 
the Miami, Sauk and Fox, Kickapoo, Menomini, and Shawnee 
agree among themselves and differ from the Ojibwa in classing 
the mother's brother's son and father's sister's son with the uncle 
and nephew, respectively; the corresponding female cousins with 
the mother and daughter, respectively.^ But this is precisely like 
the system of the Winnebago and Southern Siouans, as Morgan 
himself expressly states! Why do the Algonquian tei*minologies 
cited resemble those of a group of Siouan tribes rather than the 
Ojibwa nomenclature? 

In order to appreciate the significance of the actual facts, we 
must plot them on some such map as that of Mooney in his paper 
"The Cheyenne Indians." ^ We then find that the system in ques- 

* R. H. Lowie, "Exogamy and the Classificatory Systems of Relationship," 
American Anthropologist, XVII (1915), 237 f,, reprinted as No. 2 in tlie present 

''Morgan, op. cit., pp. 211-217. My attention was drawn to these facts by 
Mr. Leslie Spier. 

* James Mooney, The Cheyenne Indians, AAA, Memoirs, Vol. I, Part VI (1907). 


Interpretations of Kinship Terminologies 

tion is spread over an absolutely continuous area, covering the ter- 
ritory of the Menomini, Winnebago, Iowa, Omaha, Ponca, Oto, 
Kansas, Osage, Illinois, Sauk and Fox, Miami, and Shawnee. But 
the Ojibwa also form part of a fairly continuous area, — that in- 
cluding besides their own habitat that of the Cree, Dakota, 
Wyandot, and Iroquois, all of whom designate cross-cousins as 
cousins. Since within both areas members of distinct linguistic 
families share the same kinship features, the similarities, unless 
explicable by similar social conditions, can be explained only by 

The similarities cannot be explained by similar social condi- 
tions. The Ojibwa do not possess either the clan or the moiety 
system of the Iroquois, yet they have a similar nomenclature. 
They do possess the gentile organization of the Central Algon- 
quian, yet that similarity was not an adequate cause for the 
production or maintenance of a system similar to that of the Cen- 
tral Algonquians. No sociological condition can be conceived 
that might account for the empirical distribution of traits. On the 
other hand, that distribution corresponds so closely to the cri- 
terion ordinarily demanded for a proof of diffusion that diffu- 
sion, and diffusion only, must be accepted as the explanatory 

For the present discussion the distribution of systems with a 
clear development of reciprocal terms is important. To choose 
a common example, grandparent and grandchild are designated 
in these systems by a common term, or at least by a common root 
to which a diminutive affix is attached to distinguish the junior 
relative. So far as I know, systems in which reciprocal terms 
figure at all conspicuously are completely lacking in the Eastern 
Woodland, Southeastern, and Plains areas. On the other hand, 
they have been recorded among the Lillooet, Squawmish, Okina- 
gan, Spokane, and Nez Perce; ^ the Wishram, Takelma, Uintah 
Ute, and Kaibab Paiute; ^^ the San Carlos Apaches; ^^ the Nav- 

" For the first three, only so far as great-grandparent and great-grandcliild are 
concerned. Franz Boas, "Terms of Relationship of the Salish Languages," Report 
of the Sixtieth Meeting of the British Association for the Advancement of Science, 
1890, pp. 688-692; Morgan, op. cit., pp. 245, 249. 

'^° Edward Sapir, "A Note on Reciprocal Terms of Relationship," American An- 
thropologist, XV (1913), 132-138. 

" Oral communication by Dr. Goddard. 


Kinship and Social Organization 

aho; ^^ the Zuni; ^^ the Tewa, Acoma, and Cochiti; ^^ the Pa- 
pago; ^^ the Yokuts; ^^ the Kern River and Mono; ^^ the Paviotso, 
Moapa and Shivwits Paiute, Wind River and Lemhi Shoshone; ^^ 
and the Kootenay.^^ Having regard to our very meager informa- 
tion on some of the tribes concerned, everyone must again be 
impressed with the fact that the trait discussed is spread over a 
large, practically continuous area, among more than half a dozen 
distinct linguistic families, while outside that area it is lacking. 

This phenomenon of distribution cannot be explained soci- 
ologically. The tribes cited vary fundamentally in social organiza- 
tion and usage. Nothing could be more distinct in this regard 
than the clan division of the Pueblo tribes and the loose organiza- 
tion of the Plateau Shoshoneans. Moreover, no social usage that 
could conceivably unite grandparent and grandchild has ever been 
reported in North America, no trace of anything like the Australian 
class system being known. 

A more minute examination supports the theory of diffusion. 
Sapir has shown that "in this matter of relationship terais two 
such closely related dialects as Ute and Southern Paiute dififer 
on a point on which they respectively agree with a neighboring 
Shoshonean and with a non-Shoshonean language ( Tewa ) . Here, 
as often, a cultural dividing line runs clear across a homogeneous 
linguistic group." ^° Similarly, Mr. Gifford informs me that the 
Kern River people share with the Southern Paiute the use of 
diminutive sufiBxes for the junior relative designated by a re- 
ciprocal term. These tribes, though belonging to distinct branches 
of the Shoshonean family, are geographically contiguous. Why 

" Franciscan Fathers, An Ethnologic Dictionary of the Navaho Language ( St. 
Michaels, Ariz., 1910), pp. 435-436; personal communication by Dr. Elsie Clews 

" Communication by Dr. Kroeber. 

^* Communications by Drs. Parsons and Radin; J. P. Harrington, "Tewa Relation- 
ship Terms," American Anthropologist, XIV (1912), 472-498. 

'" Dr. Kroeber. 

" A. L. Kroeber, The Yokuts Language of South Central California, Univ. CaUf. 
Publ. Amer. Arch, and Etlin., II (1907), 240. 

" Mr. GifFord. 

^ My own field notes. 

" Professor Boas. 

^ Sapir, loc. cit. 


Interpretations of Kinship Terminologies 

they should agree in a feature not common to the entire family 
is not at all clear unless we assume that the similarity is a contact 
phenomenon. Here again a sociological interpretation is barred at 
the outset: what social usage or institution could lead to the em- 
ployment of a diminutive suffix? 

I can refer only briefly to certain other points. The occurrence 
of the distinction between vocative and non-vocative forms is one 
of the phenomena that should be investigated and plotted on a 
map. It is markedly developed among Siouan tribes, but has also 
been noted by Uhlenbeck for the Blackfoot,^^ by Morgan for the 
Nez Perce and Yakima,^^ by Sapir for the Wishram. Before adopt- 
ing any interpretation we must determine more definitely the 
distribution of this trait. 

Another very interesting feature is the differentiation of ma- 
ternal and paternal grandparents. This is so completely lacking 
in the immense area east of the Rocky Mountains that Morgan has 
not even distinct tables for these relationships and notes their 
discrimination with some surprise for the Spokane. ^^ Yet in the 
Far West of the United States it is exceedingly common. Boas 
notes it for the Kalispelm and Okinagan,-^ it occurs probably 
among the majority of Plateau Shoshoneans,"^ among the Takelma 
and Wishram,-*' and according to Professor Kroeber is widespread 
in California. According to Harrington and Kroeber, it exists also 
among the Tewa and Zufii; weakly developed, it occurs among 
the Navaho. It is a striking fact that in the North American areas 
in which clans, gentes, and moieties occur (including the Tlingit 
and Haida ) and where accordingly a discrimination between ma- 
ternal and paternal grandparents might a priori be expected, such 
a distinction exists only in the Southwest, while the discrimination 
occurs precisely in the region without definite social organization. 
We are clearly dealing with an historical problem. 

Still another peculiarity may be mentioned here because its 

^ C. C. uhlenbeck, "Exogamy of Peigans," Internationales Archiv fur Ethno- 
graphic, XX (1912), 205. 
^ Morgan, op. cit., pp. 249 f. 
=" Ibid., p. 247. 
^ Boas, loc. cit. 

'^' Sapir, loc. cit., and my own field notes. 
"" Sapir, loc. cit. 


Kinship and Social Organization 

occurrence seems restricted to the same general area — the change 
of terms after the death of a near or connecting relative. Tliis 
has been found by Professor Kroeber among the Yokuts and by 
Mr. Gifford among the neighboring Kern River and Kawaiisu 
Shoshoneans of California; Boas recorded it for several Salish 
tribes, the Chinook, and the Kootenay.^^ Again one cannot think 
of a social custom prevalent among these tribes capable of pro- 
ducing the terminological feature and lacking among tribes 
without it. 

Finally, I may refer to the absence of separate terms for elder 
and younger brothers and sisters. The failure to distinguish these 
diflFerences within one's own generation marks the Pawnee sys- 
tem,^^ and occurs in that of the Kiowa.^^ The fact that so glar- 
ing an anomaly from the point of view of the ordinary North 
American system should occur in the same portion of the Plains 
area is hardly without significance, and, once more, we are 
tempted to ask what social usage can be lacking here that unites 
the rest of the North American tribes. 

The foregoing remarks seem to me to establish the principle 
that features of kinship terminology are distributed like other 
ethnographical phenomena and must be approached in the same 
spirit. Like specific customs, beliefs, or implements, particular fea- 
tures of kinship nomenclature are sign-posts of cultural relation- 
ship. This being so, it becomes obvious that in order to get the full 
benefit of relationship systems for historical purposes it does 
not sufiice to record the fundamental elements of a terminology. 
On the contrary, as in other cases apparently trifling features are 
the most important because when they are found to occur in 
distinct tribes a specific liistorical connection is indicated. It is 
likewise clear that after such a general qualitative orientation as 
I have given above, far more detailed and in some measure quanti- 
tative studies along lines suggested by Professor Kroeber must 
set in. It is not the same thing whether half the terminology of a 
tribe is reciprocal or whether reciprocity is expressed only for 
the paternal grandfather and the son's son; nor whether classifica- 

^' Boas, loc. cit.; idem, "The Vocabulary of the Chinook Language," American 
Anthropologist, VI (1904), 135; and manuscript notes. 
"^ Morgan, op. cit., p. 197. 
"" My own field notes. 


Interpretations of Kinship Terminologies 

tory terms (in an etymological sense) exist like our own *^'cousin'* 
and "uncle" or whether the classification includes father's brothers 
with the father, as in many primitive systems. A much finer sta- 
tistical treatment is likely to reveal the accentuation of traits in 
a particular center of distribution and their gradual diminution 
as we pass toward the periphery of the area in question. If, as 
appears rather likely from certain indications, the entire western 
slope of the United States should then be found to differ markedly 
from the rest of the continent in point of kinship terminology, 
that result would be of considerable historical significance, over 
and above the value of particular historical inferences. 

It remains to say a word on the relation of the historical to 
the sociological point of view. In a paper already cited I have 
expressed my belief in the validity of sociological interpretation, 
and since I still adhere to this conviction I feel it incumbent to 
harmonize with it the apparently contradictory argument of 
the preceding pages. 

As stated above, I am in favor of dealing with kinship terms in 
the same way as with other ethnographic features. In the in- 
terpretation of cultural resemblances the criterion for historical 
connection generally applied by American investigators is oc- 
currence in a continuous area or among tribes of known historical 
relations, while in other instances they assume independent de- 
velopment. This procedure does not solve all perplexities but 
seems the only one that can yield any assistance at all, and it is 
applicable in exactly the same way to the problems at hand. 
When I find a reciprocal system among the Yokuts and Kern River 
Indians, I explain the similarity by diffusion; when I find a recipro- 
cal system among the Australian Arunta and the Kern River In- 
dians, I do not. In this latter case the question naturally arises 
how the similarity could arise independently. If the Califoniians 
possessed a class system like that of the Arunta, I should not hesi- 
tate to ascribe the similarity to this institution; since neither they 
nor any other American tribes have anything of the sort, I con- 
fess that I have no explanation to offer. But we are not alwa\'s so 
unfortunate. The fact that the Tlingit and Crow class the father's 
sister's daughter with the father's sister is intelligible from the 
common possession of a maternal exogamous grouping by these 
tribes. Here as among the Choctaw, certain Pueblo and some 


Kinship and Social Organization 

Melanesian tribes the clan factor has proved stronger as a classifi- 
catory device than the generation factor. Obviously clan exogamy 
does not furnish a complete explanation, since it does not produce 
the same effect among the Iroquois, to cite but one illustration. 
Whether we can adduce an additional sociological determinant, 
remains to be seen. At all events, we may say that the trait in \ 
question is a function of matrilineal descent plus certain unknown , 
factors tending to an accentuation of the rule of descent. So, 
wherever we can comiect empirical resemblances between un-' 
related tribes with actual social customs from which those re- 
semblances naturally flow, we have an adequate explanation of 
the phenomenon of similarity and do not require recourse to bor- 
rowing or cultural relationship. 

As everywhere, so here there will be room for doubt and sub- 
jective interpretation. For example, who would state categorically 
that the clan system of the Muskhogeans did or did not give rise 
to the merging of collateral and lineal kin independently of the 
development of this feature among the other tribes of the Eastern 
clan area? Here we find both similar conditions adequate to the 
production of the terminological trait and contiguity of territory. 
I can conceive the diflfusion of a clan organization with a cor- 
related terminology at a relatively early period over the entire 
Eastern area; but I can also conceive an independent develop- 
ment of the tenninology from a mere diffusion of the clan concept 
(not to consider other logical possibilities). If the Tlingit and 
Haida developed a "classificatory" system independently, so could 
the Choctaw or the Iroquois. In such instances I tliink it best 
to admit frankly either that we can not make up our minds or that 
our theoretical preferences rest on more or less subjective grounds. 
In this regard the study of kinship nomenclature does not differ 
in the least from like investigations of other cultural elements. 
And I hope the foregoing remarks have illustrated the axiom, 
I sometimes ignored by ethnologists, that every ethnological prob- 
lem is primarily a problem of distribution. ' 



The Kinship Systems of 
the Crow and Hidatsa 

In two previous publications ^ I have expressed Ml 

agreement with the view that kinship terminologies are correlated 
with social organization and have advanced North American data 
in defense of this proposition. I fully realize that in pointing out 
this correlation I have made merely a beginning. If the correla- 
tion is directly with exogamy, we are confronted with the diffi- 
culty that sometimes we do not know which of the two units — 
say, the Iroquois moiety and the Iroquois clan — is the primarily 
exogamous division. If, on the other hand, as I hinted in my shorter 
paper, the correlation is directly with divisions tracing descent 
in a definite way and only indirectly with exogamy, correspond- 
ing perplexities await us. Is the Hidatsa system, for instance, as- 
sociated with the non-exogamous moieties of the exogamous 

Proceedings of the Nineteenth International Congress of Americanists, Wash- 
ington, December 27-31, 1915 (Washington, 1917), pp. 340-343. 

^ R. H. Lowie, "Exogamy and the Classificatory Systems of Relationship," Amer- 
ican Anthropologist, XVII (1915), 223-239, and in Proceedings of the National 
Academy of Sciences, I ( 1915), 346-349; reprinted as No. 2 in the present volume. 


Kinship and Social Organization 

clans, or have both factors jointly molded the nomenclature of 
this tribe? Similar questions arise among various tribes, for ex- 
ample among the Central Algonkians and in the Southeastern 
culture area. 

The foregoing remarks rest on the assumption that social causes 
affect the kinship terminology. While, however, I accept this part 
of Dr. Rivers' theory, I believe his proposition that all features 
of kinship nomenclature are based on social conditions is without 
foundation, as I hope to demonstrate in the present paper. 

In order not to prejudice the matter, let us assume that any given 
system of relationship is the result of an indefinite number of 
conditions. Then we can disentangle particular determining con- 
ditions by preserving the rest identical or as nearly identical as 
possible. Let us, then, take two tribes as closely allied as possible 
in history, culture, and language; and let us seek to correlate cul- 
tural differences of whatever kind occur with the observed dif- 
ferences in nomenclature. 

The Crow and Hidatsa seem to furnish an ideal test-case. Their 
languages are very closely related, and the clan concept may be 
said to be identical. While the cultural differences are appreciable, 
the total absence of such is not to be expected on Dr. Rivers' 
theory, since otherwise the differences in terminology would re- 
main without cause. Let us determine, then, to what extent ob- 
served terminological differences are traceable to a difference in 
social usage. 

Before proceeding to a consideration of the differences it will 
be useful to summarize the most significant features the two sys- 
tems possess in common. In the first place, both belong to the 
widespread centripetal, or, as it is more commonly called, "classifi- 
catory" type — the type in which collateral relatives are not dis- 
tinguished from lineal relatives. Secondly, both differ from many 
systems of this type in lacking any term for uncle, cousin, nephew, 
and niece, the existing term for aunt (father's sister) being thus 
without a correlative. Thirdly, in connection with the peculiarity 
just mentioned, there is a merging of distinct generations, (a) 
The mother's brother is classed with the elder brother and ad- 
dresses his sister's son and daughter as his younger brother and 
sister, respectively. This is an anomaly already noted by Morgan, 


Kinship Systems of the Crow and Hidatsa 

who found nothing Hke it among other Indian tribes.^ (h) Cross- 
cousins are not cousins as among the Iroquois and Dakota, but 
belong to distinct generations, the mother's brother's son and 
daughter being classed with the son and daughter; the father's 
sister's son with the father; and the father's sister's daughter (as 
well as all her female descendants through females) with the 
father's sister.^ The cross-cousin terminology of the Crow and 
Hidatsa is not confined to them. The Tewa of Hano call the 
father's sister's son "father." * Among the Hopi the same usage 
prevails, and in addition the father's sister's daughter is grouped 
with the father's sister and the mother's brother's son with the 
son.^ The Tlingit have a single term for the father's sister and the 
father's sister's daughter, and a single term for the father's brother 
and the father's sister's son.*^ Morgan himself cites the Muskogean 
and the Pawnee in this connection. As a representative of the 
former group the Choctaw are said to class the father's sister's son 
with the father, and ( man speaking ) the father's sister's daughter 
with the father's sister, while the mother's brother's son and 
daughter are classed with the son and daughter.'^ The Pawnee 
class the mother's brother's son and daughter with the son and 
daughter, respectively; the father's sister's son with the father; 
and the father's sister's daughter with the father's sister, who is 
classed with the mother as in the Crow vocative form, but unlike 
the Crow non-vocative and the Hidatsa common form.^ The oc- 
currence of a centripetal nomenclature with disregard of genera- 
tions in the designation of cross-cousins and the distinctive group- 
ing of the maternal uncle with the elder brother sufficiently 
characterize the Crow-Hidatsa system for the present purpose. 
Let us now turn to the more striking diflFerences in the nomen- 
clature of the two tribes. 

^ L. H. Morgan, Systems of Consanguinity and Affinity (Washington, 1871), 
pp. 188-189. 

^ For the Crow the last statement appHes only in non-vocative usage. 

* W. H. R. Rivers, Kinship and Social Organisation (London, 1914), pp. 53-54. 

^ My own field notes. 

' J. R. Swanton, "Social Condition, Beliefs, and Linguistic Relationship of the 
Tlingit Indians," BAE, 26th Ann. Kept. ( 1908), pp. 424-425. 

'' Morgan, op. cit., p. 191. It should be noted that the data associated with those 
cited above are anomalous and that corroborative testimony is highly desirable. 

« Ibid., pp. 197-198. 


Kinship and Social Organization 

Taking first the blood relatives of one's own generation, the 
Hidatsa have seven terms distinguishing the age of the speaker 
relatively to the person addressed or spoken of, while the Crow 
have only six. That is, while Crow males and females use a com- 
mon term for the elder sister, masaka'ata, the Hidatsa have two 
terms, matawVa, used by men, and maru'u, used by women. It is 
not easy to conceive a social reason for such a difference, but in 
this case we are fortunately not compelled to use purely negative 
evidence. The Crow term is not connected with either of the 
two Hidatsa roots or with any other Hidatsa stem of approximately 
the same meaning. It is, however, fairly clearly related to the non- 
vocative Crow designation for "my mother," masake, from 
which it seems to be formed by fusion with the diminutive end- 
ing -ka'Ha. On the other hand, the Hidatsa matawVa is the exact 
phonetic equivalent (having regard to the existing linguistic re- 
lations of the two languages) of the Crow bacbVa. Stripped of 
the initial possessives, we have here a common stem, bVa, 
"woman," which in intervocalic position becomes wVa. In Crow 
bacbVa is a generic term for "my sister" or "my clan sister," re- 
gardless of age; in Hidatsa the equivalent expression has the 
specialized meaning of "my elder sister" (m. sp.). We are clearly 
dealing with a purely linguistic phenomenon: either the Crow 
have generalized the meaning of a stem that formerly had a more 
restricted meaning, or the Hidatsa have come to attach a narrower 
significance to the common stem. German Hund compared with 
English hound, and German Dogge compared with English dog 
are familiar examples of these processes. 

In the first ascending generation of blood-kin there are several 
noteworthy differences. While the terms for father are coexten- 
sive, the Crow have distinct terms for men and women speak- 
ing, while the Hidatsa use a common word. It is not easy to sug- 
gest a sociological explanation. Again, while the stems for mother 
(vocative) are almost identical and nearly coextensive in mean- 
ing, the Crow word is applied to the father's sister for which there 
is a quite distinct expression in non-vocative usage, correspond- 
ing to an Hidatsa term that is used both in direct address and 
non-vocatively. What sociological reason could conceivably be 
adduced to account for the classification of the same relative with 


Kinship Systems of the Crow and Hidatsa 

the mother in direct address and as distinct from the mother in 
non-vocative terminology? Comparison not only with Hidatsa 
but with Siouan nomenclatures generally seems to show clearly 
that we are here dealing with an anomalous extension of mean- 
ing, a word originally restricted to the mother, her sisters, and 
her cousins on the maternal side having come to be applied to 
the father's sister in direct address. Another curious usage may 
be mentioned before leaving this generation. In a number of 
cases the Crow distinguish non-vocative from vocative endings 
not by a change of stem but by altering terminal -a to -e. Now, 
while a father's brother or clansman is addressed axe like the 
father, the non-vocative is ma'sake, and for all ordinary purposes 
the correlative vocative form ending in -a is lacking. Neverthe- 
less, the term exists, but it is restricted exclusively to prayers to 
the Sun, in which case the significance given is "paternal uncle." 
Here once more we are obviously dealing with a linguistic phe- 

In the descending generations the most impressive difference 
relates to the designation of the grandchild. The Hidatsa use a 
common stem for the vocative and non-vocative, the former 
being mdtawapicd, which receives a slightly different ending in 
the non-vocative. The Crow have the exact phonetic equivalent, 
macba^pite, but it is used only in the non-vocative; the grandson 
is uniformly addressed as "son," the granddaughter as "daughter." 
Of the facts themselves there can be no doubt; in addition to other 
evidence there is that of an important myth secured in text in 
which the hero, though named Old-Woman's-Grandchild, is per- 
sistently addressed as "son" by his adoptive grandmother. The 
grandson does not enter a new social category when he is ad- 
dressed that makes him similar to the son and different from 
himself when merely spoken of. We are dealing, then, once more 
with a capricious phenomenon of language. 

Turning to terms of affinity, we find one difference in ter- 
minology that in a certain sense is connected with a difference 
in social usage. A Crow man never speaks to his wife's brother's 
wife; accordingly, there is no vocative for this relative and non- 
vocatively she is grouped with the other persons with whom social 
intercourse is tabooed. Among the Hidatsa no such taboo exists 


Kinship and Social Organization 

for this aJBBnity and consequently she may be addressed and is 
not classed with tabooed aflBnities but with the daughter-in-law 
and correlatively addresses her husband's sister's husband by the 
common term for father-in-law (w. sp.) and grandfather (m. 
sp., w. sp.). In this classification an influence of social organiza- 
tion is conceivable, inasmuch as the Hidatsa clans are grouped 
together in non-exogamous matrilineal moieties; a man would 
therefore be in the same moiety as both his wife's brother's wife 
and his son's wife, provided the moiety was once exogamous. 
For this, however, evidence is wanting. On the other hand, an- 
other difference cannot be thus accounted for. The father's sister's 
husband is classed by the Crow with the father, though he can- 
not possibly be in the same clan as the father; a plausible explana- 
tion is that this classification results from the designation for the 
father's sister, who is vocatively called "mother." We are dealing 
with a primarily linguistic phenomenon, then, as explained above. 
Among the Hidatsa the father's sister's husband (m. sp., w. sp.) 
is classed with the father-in-law (w. sp.). So far there is no 
particular difficulty beyond assuming exogamous moieties, since 
these relatives both would then belong to the same moiety. But 
the term that embraces these relatives is also applied indiscrimi- 
nately to the maternal and the paternal grandfather (m. sp., w. 
sp. ), who cannot possibly be grouped together on sociological 
grounds if we retain the basic hypothesis. It might be suggested 
that the term was originally restricted to the father's father, but 
this would be not only gratuitous assumption but contrary to all 
known Siouan usage, whether Omaha, Winnebago, Dakota, or 
Crow. The fact that the Hidatsa place the father's sister's husband 
in the same category with blood-relatives of the second ascend- 
ing generation while the Crow class him with the first ascending 
generation is therefore not satisfactorily explained by the one 
important feature of social organization in which the tribes 

It seems unnecessary to pursue the comparison in greater de- 
tail. The most important observed sociological difference between 
the Hidatsa and Crow — the presence of moieties among the 
former — is either not correlated with the terminological differ- 
ences at all, or it is correlated with them in a subordinate way, 


Kinship Systems of the Crow and Hidatsa 

and then only if we make the auxihary hypothesis, unsupported 
by any evidence, that the moieties were originally exogamous. 
On the other hand, the observed differences in nomenclature 
are of exactly the same type that we constantly have occasion 
to note in closely related languages. Thus, to turn to another por- 
tion of the Crow and Hidatsa vocabularies, the Crow word lEXese 
denotes a snake generically, but in Hidatsa it is applied to a par- 
ticular species, while the Hidatsa word for snake means in Crow 
either a wasp or an otter. The same specialization and exten- 
sion of significance has taken place in the kinship terminology. 
If we must recognize on evidence not presented in this paper a 
far-reaching influence of social conditions on kinship nomencla- 
ture, we must admit with equal decisiveness that many of its 
features are irreducible to a social basis and represent exclusively 
linguistic phenomena. 




Family and Sib 

Ethnologists in the United States are agreed that 
the North American peoples of crudest culture are loosely or- 
ganized, with the family as the basic unit; that tribes definitely 
organized into sibs (Morgan's gentes, clans of English writers) 
represent a higher cultural plane at which, however, the influence 
of the family is clearly discernible; that accordingly the sib is a 
later, superimposed product, not the invariable predecessor of 
the family. It remains to define the mechanism by which such a 
transformation might have been effected. 

The sib, like the family, is a kinship group. It is at once more 
and less inclusive than the rival unit. On the one hand, it ex- 
cludes one half of the blood-kindred — the father's side of the 
family in matronymic, the mother's side in patronymic societies. 
On the other hand, it admits on equal terms all kindred of the 
favored side regardless of degree and even individuals considered 
blood-relatives merely through legal fiction, whence the rule of 
sib exogamy. The sib normally embraces not merely the de- 
scendants through females of an ancestress, or through males of 
an ancestor, but several distinct lines of descent, which are only 
theoretically conceived as a single line. This particular form of 

American Anthropologist, XXI (January-March, 1919), 28-40. 


Family and Sib 

inclusiveness, based on adoption, coalescence of ceremonial units, 
or what not, is too familiar a phenomenon to present any great 
difficulty to our comprehension. The real problem lies in the 
origin of what Dr. Goldenweiser calls the maternal and the pater- 
nal family pattern rather than in the expansion of these unilateral 
bodies of kindred to form larger groups of the same type and in 
theory identical with them. 

It is my purpose to show that the characteristic features j,^^-- 
of the sib organization are in some measure prefigured among 
sibless tribes; that certain usages may bring about an alignment 
of kin such as occurs in sib systems; that the sib is in fact merely 
a group of kindred thus segregated and defined by a distinctive 

In the interests of clearness it is well to define at the outset 
the relation of my present position to that assumed in previous 
publications.^ Elsewhere I argued that the "Dakota" principle 
of classifying kin is logically and actually associated with sib 
systems and lacking in sibless tribes. Accordingly I concluded 
that the sib was the antecedent condition for the development 
of the Dakota type of relationship nomenclature. At present I 
should say that while the empirical correlation holds true the 
causal relations are to be reversed; generally speaking, a particu- 
lar grouping of kin resulted in a sib system, though a fully es- 
tablished sib organization can and did in turn influence the 
nomenclature of kin. 

In comparing the nomenclatures of sibless and of definitely 
organized tribes, we often find two characteristic differences. [^^^' 
The former either fail to distinguish paternal and maternal rela- 
tives or they fail to merge collateral and lineal kin, or both. For 
example, the Coast Salish have a single term for paternal and 
maternal uncles, but distinguish children from all nephews and 
nieces. However, the terminologies of these peoples are by no 
means uniform and in many of them we can detect foreshadow- 
ings of the Dakota principle. 

The most obvious of these is the classification not merely of 

^ R. H. Lowie, "Exogamy and the Classificatory Systems of Relationship," Amer- 
ican Anthropologist, XVII (1915), 223-239 (reprinted as No. 2 in tlie present 
volume); idem, Culture and Ethnology ( New York, 1917 ), chap. v. 


Kinship and Social Organization 

kindred but of unrelated tribesmen as well, nay sometimes even 
of strangers, according to age. Dr. Karl von den Steinen was 
called elder brother by the Bakairi, maternal uncle by the Me- 
hinaku.^ That is to say, approximate age-mates are classed together 
except so far as they are diflPerentiated by sex. This principle may 
be designated as Hawaiian, since it is most consistently followed 
by the Hawaiians and related Polynesians and Micronesians. 
Elsewhere, however, we do find suggestions of Hawaiian classifi- 
cation among loosely organized peoples. Perhaps the most common 
extension occurs in the second ascending generation, any vener- 
able individual being addressed as a grandparent. To cite non- 
American examples, this is recorded for the Hottentot,^ and the 
Chukchi draw no distinction between grandfather and great- 
uncle, grandmother and great-aunt."* The Chukchi nomenclature 
reveals other approximations to the Hawaiian pattern. There is 
no distinction between maternal and paternal uncles or aunts, 
and even those once removed are designated by the same terms. 
On the other hand, the Chukchi differ fundamentally from tribes 
following either the Hawaiian or the Dakota plan in rigidly 
separating the father from all uncles, the mother from all aunts. 
In North America there are interesting analogies. The Wind 
River Shoshoni, I found, class all cousins with brothers and sis- 
ters, conforming to tliat extent wholly to the Hawaiian scheme; 
and Sapir notes the same feature for the Nootka. With the Hupa 
all women of the second ascending generation are grandmothers, 
all the old men grandfathers, all the children born in the same 
house one another's siblings.^ The Coast Salish go at least equally 
far. Here not only are great-uncles and grandfathers classed to- 
gether and reciprocally call their own and their siblings' grand- 
children by a common term, but all cousins are grouped with 
brothers and sisters, while a single term denotes father's and 
mother's siblings. One step further and in the first ascending 
generation, too, they would follow the Hawaiian principle; tlie 

^ Karl von den Steinen, Unter den Naturvolkern Zentral-Brasiliens, 2te Aufl. 
(Berlin, 1897), p. 286. 

^ L. Schultze, Aus Namaland und Kalahari (Jena, 1907), p. 300. 

* Waldemar Bogoras, The Chukchee, AMNH, Memoirs, XI (Leiden, 1909), p. 538. 

° P. E. Goddard, Life and Culture of the Hupa, Univ. CaUf . Publ. Amer. Arch. 
andEthn., I (1903), 58. 


Family and Sib 

step, however, is not taken since uncles and aunts remain dif- 
ferentiated from parents.® 

Such extensions of terms as have been cited hardly require 
special psychological explanation since they are not unfamiliar 
among ourselves. Among primitive tribes there exists the ad- 
ditional stimulus of a widespread and intense aversion to the use of 
personal names. But the tendency to designate individuals by 
a common term may have far greater than merely terminological 
significance. Because primitive peoples attach an extraordinary 
importance to names the more remote cousin who is called cousin 
or sister may become more closely related in thought and mar- 
riage may be tabooed regardless of degree of propinquity. This 
we are specifically told in the case of the Paviotso.''^ Among the 
Nez Perce even third cousins were not allowed to marry ^ and 
the union of second cousins roused ridicule in Thompson River 
communities.^ I conjecture that these are analogous cases. 

However, the merging of remote and near collateral kin, or 
even of collateral and lineal lines of descent, does not suflBce to 
pave the way for a sib organization; in addition to inclusiveness 
there must be dichotomy, that is, the extension must be unilateral 
not Hawaiian. Although our knowledge of the social organization 
of sibless tribes remains sadly inadequate, a number of cases 
can be presented in which there is definite bifurcation of blood- 
kindred. For the present a few illustrations must suffice; they are 
selected from four tribes typical of the great sibless area and 
representing distinct linguistic stocks. 

Chinook lo 

ma'ma, -ma, am, father -naa, -a, mother 

-motx, father's brother -klotcxa, mother's sister 

-ta, mother's brother -lak, father's sister 

" Franz Boas, "Terms of Relationship of the Salish Language," Report of the 
Sixtieth Meeting of the British Association for the Advancement of Science, 1890, 
pp. 688 ff. 

"^ S. Hopkins, Life among the Piutes (Boston, 1883), p. 45. 

* H. J. Spinden, The Nez Perce Indians, AAA, Memoirs, II, Part III ( 1908), 250. 

° James Teit, The Thompson Indians of British Cohimbia, AMNH, Memoirs, I 
(1900), 325. 

^° Franz Boas, "The Vocabulary of the Chinook Language," American Anthro- 
pologist, VI (1904), 135. 


Kinship and Social Organization 

na, father 

hai'i, father's brother 

atsi, mother's brother 

Paviotso 1^ 

Porno 12 

e, harik, father 
keh, father's brother 
tsets, mother's brother 

Okanagan ^^ 

lEe'u (m. sp.), father 
mistm (w. sp.), father 
sm'e'elt, father's brother 
sislf, mother's brother 

pia, mother 

pidu'u, mother's sister 

pahwa, father's sister 

te, nik, mother 
tuts, mother's elder sister 
sheh, mother's younger sister 
tveh, father's sister 

sk'oH (m. sp.), mother 
torn (w. sp.), mother 
swdwa'sd, mother's sister 
sk'o'koi, father's sister 

Such dichotomy of kin as is here indicated is exactly what might 
be expected under that family organization which American stu- 
dents regard as prior to a sib system, for since the parents belong 
to different families their relatives are logically enough distin- 
guished from one another. 

Let us now assume that the bifurcating and the merging tend- 
ency as hitherto expounded unite. Then we shall have a ter- 
minology in which all the mother's female kindred belonging 
to her generation will be classed with the mother's sister, all of 
her male kindred in that generation are treated as mother's 
brothers, while corresponding classification is given to the father's 
relatives. In that generation we shall have an alignment anticipat- 
ing that of the Dakota type, from which it diflPers solely in the 
distinction maintained between parent and parent's sibling of the 
same sex. 

What happens, however, in the speaker's generation? Corre- 
sponding to the four uncle-aunt terms we might logically expect 
an equal number of cousin terms, or even twice as many tlirough 
sex discrimination. As a matter of fact, the classification of 
cousins follows quite diflPerent principles. In some nomenclatures 

^^ A. L. Kroeber, California Kinship Systems, Univ. Calif. Publ. Amer. Arch, 
and Ethn., XII (1917), 359. 
^ Ibid., pp. 370 f. 
" Boas, in "Terms of Relationship of the Salish Language," p. 691. 


Family and Sih 

of sibless tribes, e.g., the Paviotso and Shoshoni, the Hawaiian 
principle is apphed and all cousins are brothers and sisters. Among 
the Coast Salish we find the same grouping but also a specific 
term for cousin. I assume — and this is the most hypothetical fea- 
ture of my scheme — that at the stage preceding the evolution of 
the sib the natives had specific terms for brother and sister, while 
all other relatives of that generation were lumped together under 
a single term except so far as they were differentiated according 
to sex. This would yield a grouping somewhat similar to that 
in the first ascending generation since the members of the im- 
mediate family would be segregated from more remote kin. On 
the other hand, this classification would differ from that char- 
acteristic of most tribes with a sib organization. For one of the 
essential features of their nomenclatures lies in the dichotomy of 
cousins according to the likeness or unlikeness of the sex of the 
parents through whom the relationship is established. In per- 
haps the most common variety of the Dakota scheme parallel 
cousins are brothers and sisters, cross-cousins are designated by a 
distinct cousin term. 

It is essential to point out that no perfectly satisfactory expla- 
nation of this classification has been given except on Tylor's hy- 
pothesis that it originated in a moiety organization.^^ The hy- 
pothesis that parallel cousins are simply moiety mates admirably 
accounts for the grouping but does not cover the facts of distri- 
bution, since the division into parallel and cross-cousins is often 
found with a multiple sib system. ^^ This, however, in turn fails 
to account for the classification. If there are only two sibs in a 
tribe (or, prior to sibs, only two intermarrying families), cross- 
cousins are in one moiety and parallel cousins in the other, as 
Tylor pointed out. But if there are five, the condition is very dif- 
ferent. Assuming maternal descent, the children of sisters will 
indeed belong to the same social unit but the children of brotliers 
need not; one may marry into group b, the other into group c 
and their children will belong to their respective mothers' sibs. 

" E. B. Tylor, "On a Method of Investigating the Development of Institutions," 
RAI, Journal, XVIII (1889), 264. 

^ Cross-cousin marriage, which seems closely connected with a dual organization, 
also has a distribution far too limited to account for the data. 


Kinship and Social Organization 

Now I assume that upon tribes bifurcating but merging rela- 
tives unilaterally in the manner described above, there are super- 
imposed two extremely widespread customs, the levirate and the 
sororate. The terminological effects of these usages have been 
amply discussed by Sapir,^^ though not quite adequately as regards 
cousin nomenclature. One obvious result is to obliterate the dis- 
tinction between father and father's brother, mother and mother's 
sister. In short, the Chinook and other terminologies cited (pp. 
85-86) come to conform to the Dakota principle in the first as- 
cending generation. Since father's brother and mother's sister be- 
come parents, their children become siblings, which accounts 
for the grouping together of parallel cousins. But it is not clear 
why father's sister's and mother's brother's child so often remain 
undistinguished. If, however, all cousins have previously re- 
ceived a common designation on the basis of generation, being 
differentiated only from those contemporaries who form part of 
the narrow family circle, as I assume, then the effect of the levi- 
rate and sororate is to raise parallel cousins to the status of sib- 
lings, while cross-cousins remain in the general class of con- 

I offer this suggestion not as a substitute for Tylor's interpreta- 
tion but as supplementary to it; it is designed to cover those cases 
in which parallel cousins cannot be classed together as members 
of one moiety and cross-cousins of the other for the simple reason 
that no dual organization exists, either in a fully developed or 
nascent form. 

The relation of these marriage customs to social organiza- 
tion merits some additional consideration. As to their significance 
I indorse whole-heartedly Tylor's interpretation that the levirate 
reflects a matrimonial compact not between individuals but be- 
tween families; and that for lack of actual brothers more remote 
male relatives are substituted.^^ Corresponding views of course 
apply to the sororate. Wherever our data are sufficiently explicit, 
they seem to corroborate Tylor's theory. For example, the Shasta 
purchase wives and a man is aided in the transaction by his broth- 

^^ Edward Sapir, "Terms of Relationship and tlie Levirate," American Anthro- 
pologist, XVIII (1916), 327-337. 
" Tylor, op. cit., p. 253. 


Family and Sib 

ers and relatives; accordingly it is natural that they should lay 
claim to the widow. On the other hand, a widower or the hus- 
band of a barren woman might take as his second spouse one of 
his wife's unmarried sisters or cousins. ^^ Thompson River Indian 
practice closely conforms to that of the Shasta; more particularly 
a man held an incontestable claim to his brother's widow. ^® 

In a discussion of Dr. Sapir's paper on the levirate ^° I raised 
certain diflBculties, some of which would militate no less against 
the position I now assign to these usages than against Dr. Sapir's 
explanation of kinship nomenclatures. Probably the most im- 
portant of these is a chronological one: if the levirate and the 
sororate developed subsequently to the sibs they could not of 
course give rise to that classification of kin which I now regard 
as underlying the sib. Now it is true that since Tylor no one has 
taken the trouble to ascertain the precise distribution of either 
custom and his concrete data are apparently lost. But in the hght 
of my reading I am tempted to regard his result — a forty per cent 
distribution of the levirate among primitive tribes — as far below 
the figure that would be established by a count today. This seems 
certain for North America; and here we find the interesting result 
that levirate and sororate are found jointly almost throughout the 
great sibless area — among the Salish of British Columbia, in our 
Pacific states, and the Great Basin. They are thus characteristic 
of the simpler sibless cultures, but they also appear commonly 
on a higher level with the sib system. The inference is warranted 
that they are traits preceding the sib organization and in a man- 
ner preparing the way for it. 

This, to be sure, would not apply to the Pueblo area, where 
neither levirate nor sororate is in vogue. But the best-known tribes 
of this region differ rather markedly in their nomenclature from 
the Dakota norm, though in a manner not inconsistent with the 
principles I have outlined above. The Zufii group cousins of both 
sides as siblings, though applying peculiar notions in point of 
seniority wliich may here be disregarded.^^ This is quite intel- 

^ Roland B. Dixon, The Shasta, AMNH Bull., XVII (1907), 463 f. 
" Teit, op. cit., p. 325. 

^ Lowie, Culture and Ethnology, pp. 144-150. 

^A. L. Kroeber, Zufii Kin and Clan, AMNH, Anthropological Papers, XVIII, 
Part II (1917), 58. 


Kinship and Social Organization 

ligible, of course, on the principle of generations. With the Hopi 
the two kinds of cross-cousins are differentiated (see below), so 
that the problem as to their classification does not arise in the usual 
form. But what of the Zuiii and Hopi classification of uncles and 
aunts? Here, too, I can see no difficulty. Though the levirate, 
e.g., supplies an excellent specific reason for identifying father's 
brother and father while differentiating them from the mother's 
brother, the joint force of the more general bifurcation and 
generation factors is adequate to produce the same result. Since 
father's brothers thus came to be reckoned as fathers, and mother's 
sisters as mothers, the Hopi classification of parallel cousins as 
siblings follows: the cliildren of those I call my parents must be 
my brothers and sisters. 

The classification of parallel cousins, however, involves a funda- 
mental obstacle to any theory that would derive the sibs from an 
earlier system of kinship nomenclature. As Morgan himself pointed 
out, the status of sibling is not coterminous with that of sib fellow. 
In a matrilineal society only the children of sisters, not of brothers, 
belong to the same social unit, yet all parallel cousins are addressed 
as brothers and sisters."^ If we assume that the conditions de- 
scribed above gave rise to the terminology that normally ac- 
companies a sib organization, then why were some of the brothers 
and sisters taken into the sib and others discarded? 

In attempting to answer this question I desire at the outset to 
emphasize my belief in a multiple origin of the sib idea; even 
in North America I hold that there have been several centers of 
distribution. For one thing, I am strongly impressed with the 
enormous variability of the sib concept. Secondly, the generalized 
sib idea — unilateral descent — is not, as Morgan would have it, 
an abstruse quasi-metaphysical notion, but one that quite naturally 
develops from certain cultural features. These features, more- 
over, may favor either patrilineal or matrilineal descent; hence 
I see no reason why either father-sibs (gentes) or mother-sibs 
( clans ) should not have arisen directly from a loose organization 
instead of either having to evolve out of the other, though of 
course I do not reject the possibility of such a transformation. 

^L. H. Morgan, Systems of Consanguinity and Affinity ( Wasliington, 1871), 
pp. 475 f. 


Family and Sib 

To turn to the problem of parallel cousins. Sibless communities 
have often clear-cut regulations tending to establish definite lines 
of descent. The Shasta and the Thompson River Indians rec- 
ognized individual ownership of fishing stations with patrilineal 
descent of the title to them.^^ Such possessions might not loom 
large enough in the tribal consciousness to lead to significant 
consequences, they might even be outweighed by other con- 
siderations stressing the maternal lines of descent. It is quite 
diflFerent when economic privileges of some consequence are 
involved or when there is a definite rule determining the resi- 
dence of a couple after marriage, or where both these factors 
cooperate. For example, with the Bushmen, land descended in the 
paternal line; Dr. Bleek's informant occupied the site held by his 
father's father, which had descended first to his father, then to 
his elder brother, and finally to himself .^^ By such an arrangement 
sisters are separated, brothers and their descendants are united, at 
least through their property rights. In the permanent villages of 
the Hupa men were bom, Hved, and died in the same village, 
while women followed their husbands. ^^ The paternal line of vil- 
lage mates was thus inevitably stressed while the offspring of 
sisters were scattered over different localities. 

In recent years no one has emphasized the significance of such 
conditions for social organization more vigorously than Professor 
Speck. In the northeastern Algonkian region he finds non-exoga- 
mous groups transmitting hunting territories quite definitely from 
father to son and following patrilocal residence rules; brothers 
to some extent share economic privileges. ^^ Given such customs, 
it will not matter whether through the levirate and sororate all 
parallel cousins are addressed as brothers and sisters. Those paral- 
lel cousins who live together and share the same hunting preroga- 
tives, i.e., the children of brothers, will be automatically set apart 
from the children of sisters and come to be considered as in some 

^ Teit, op. cit., pp. 293 f.; Dixon, op. cit., p. 452. 

^W. H. I. Bleek and L. C. Lloyd, Specimens of Bushman Folklore (London, 
1911), pp. 305-307. 

^ Goddard, op. cit., p. 58 

^'' Frank G. Speck, "Kinship Terms and the Family Band among the Northeastern 
Algonkians," American Anthropologist, XX (1918), 143 ff.; idem. Family Hunting 
Territories, Canadian Geological Survey, Memoir 70 (Ottawa, 1916). 


Kinship and Social Organization 

respects more closely related. I regard Dr. Speck's data as most 
important in demonstrating what is to all intents and purposes a 
nascent father-sib. The external details of the processes involved 
may of course vary. For example, in the region of the northwest 
Amazons, the social unit is the exogamous house community of 
as many as two hundred individuals. Residence is patrilocal so that 
brothers take their wives to the same house. This sets up the same 
difference as among the Algonkian between the two kinds of 
parallel cousins, and here we have the interesting phenomenon 
that marriage with parallel cousins from other households, i.e., 
unions between the children of sisters, are permitted.^^ 

In considering matrilineal societies Tylor was inclined to derive 
their essential features from the basic fact of matrilocal resi- 
dence.-^ This is a luminous suggestion, for from matrilocal resi- 
dence the segregation of matrilineal kin logically follows, as does 
the exceptional status of the maternal uncle. Nevertheless a seri- 
ous obstacle to this interpretation as a general theory of the origin 
of mother-sibs lies in the restricted distribution of matrilocal 
residence even where descent is matrilineal. The Australians are 
practically all patrilocal, the Melanesians predominantly so, and 
some matronymic tribes in both Africa and America likewise have 
the wife living with her husband. There is the additional difficulty 
that residence very often is only temporarily with the wife's 
parents, in which case it suggests not infrequently merely an ob- 
ligation on the husband's part to serve for his wife in lieu or part 
payment of the bride-price. Evidently if a young couple only 
stay with the wife's parents for a year or two and then set up an 
independent household, the conditions for a matrilineal reckon- 
ing of kindred are not the same as among the Hopi or Zuni, where 
women own the houses and their husbands permanently reside 
with them. This fundamental difference between permanently and 
temporarily matrilocal residence still further restricts the applica- 
bility of Tylor's theory. Nevertheless it may be accepted as ad- 
mirably fitting the case of the Pueblo Indians, for, as Professor 

^'Thomas Whiffen, The North-West Amazons (New York, 1915), pp. 63, 66 ff. 
^ Tylor, op. cit., p. 258; idem, "The Matriarchal Family System," Nineteenth 
Century, XL (1896), 81-96. 


Family and Sib 

Kroeber has shown, the sum and substance of the Pueblo "ma- 
triarchate" hes in the female ownership of the houses. ^^ 

In attempting to supplement Tylor's explanation it seems to me 
that attention should be specially directed to economic conditions 
and the sexual differentiation of labor. Eduard Hahn has fa- 
miliarized us with the distinct character of horticulture and aratory 
culture — the former being in the hands of the women, the latter 
belonging uniformly to the masculine domain. Does not this 
suggest an interpretation of the kind required? Unfortunately we 
often lack details as to the manner of tillage, but recent data on 
the Hidatsa seem extremely suggestive. Here gardens were tilled 
jointly by the women of the maternal family and descended in the 
maternal family.^*^ That is to say, the female descendants of sis- 
ters were actually united by common property rights and associa- 
tion in economic activities. The fact that male descendants are 
not included in these labors does not seem to me fatal, for as 
soon as the joint tillers were differentiated by a name their infants 
would automatically share the same designation from birth. It is 
interesting to note that in this region there is no record of individ- 
ual hunting prerogatives of the males to counterbalance these 
horticultural privileges of the women. 

I realize that my hypothesis, even when joined to Tylor's, does 
not account for all the cases of matrilineal sibs in the world. The 
patrilocal and non-horticultural Australians and Northwest Coast 
Indians remain to be explained. Nevertheless matrilocal residence 
and the joint economic activities of women suffice to account for 
a majority of the known cases, and the residual phenomena might 
at least be approached from a similar point of view. 

I assume, then, that bifurcation and age-stratification, which 
occur among many sibless tribes, are conditions antecedent to 
the sib organization but produce an alignment of kin approximat- 
ing that of the Dakota-Iroquois nomenclatures. The levirate and 
sororate, while not indispensable, render it more probable that 
the first ascending generation should be designated after the 

-" Kroeber, Zuni Kin and Clan, pp. 47 f., 89 f. 

^Gilbert L. Wilson, Agriculture of the Hidatsa Indians (Minneapolis, 1917), 
pp. 9f., 113 f. 


Kinship and Social Organization 

normal sib fashion; and they may further bring about the usual 
grouping of cousins. But in order that sibs shall develop from 
such a terminology, it is inevitable that the children of brothers 
be differentiated from those of sisters. I follow Tylor in explaining 
part of the phenomena by patrilocal or matrilocal residence. Others 
seem intelligible from the sociological differentiation of the 
sexes and the consequent establishment of unilateral lines of de- 

When the sib has taken firm root, it is quite possible for it to 
react upon the kinship terminology. Not only may the kinship idea 
be extended to similarly named sibs of alien peoples, but the sib 
affiliation may even override the basic generation scheme, as 
among the Crow and Omaha. In these instances, too, it is de- 
sirable to view the facts in connection with associated cultural 
features. Even in such cases the terminology may sometimes re- 
sult from concrete social arrangements involved in the sib or- 
ganization rather than from the abstract concept of the sib. For ex- 
ample, the Hopi classification of the father's sister with all her 
female descendants through females simply groups under one 
head a series of house mates, which manifestly does not apply to 
the Crow or Hidatsa. 

The present is not an historical paper but a sketch intended to 
stimulate historical studies. If the sib is later than the family, we 
cannot indefinitely postpone an inquiry into the conditions that 
have moulded the sib out of a prior family organization. This in- 
volves the demand that we must learn a great deal more about 
the social life of the loosely organized peoples. The social customs 
of these tribes are no more uniform than are the sib organizations 
of other tribes. Both must be studied intensively and with con- 
stant consideration of the concomitant cultural traits if we are 
ever to frame a satisfactory theory of the development of social 



A Note on Relationship 

In his Systems of Consanguinity and Affinity, Lewis H. 
Morgan divides relationship systems into two types: 

One of these is descriptive and the other classificatory. The 
first . . . , rejecting the classification of kindred . . . describes 
collateral consanguinei, for the most part, by an augmentation or 
combination of the primary terms of relationship. . . . But the 
second . . . , rejecting descriptive phrases in every instance, and 
reducing consanguinei to great classes by a series of apparently 
arbitrary generalizations, applies the same terms to all the members 
of the same class. ^ 

As representative of the former type, Morgan cites the Aiyan, 
Semitic, and UraUan systems, wliile the American Indian (and 
primitive nomenclatures as a whole ) illustrate the second type. 

Kroeber and Rivers have criticised the basis for the distinction 
inasmuch as our English and other Indo-European temiinologies 
have such classificatory terms as "uncle" and "cousin." Morgan 
mentions these as constituting "a limited number of generaliza- 

American Anthropologist, XXX ( April-June, 1928 ) , 263-267. 

^ L. H. Morgan, Systems of Consanguinity and Affinity (Washington, 1871 ), p. 12. 


Kinship and Social Organization 

tions" but feels that their use "does not invade the principles 
of the descriptive system and that their origin lies in the con- 
stantly recurring desire to avoid the inconvenience of descrip- 
tive phrases." ^ This explanation evidently does not satisfy the 
demands of logical classification. 

Rivers also objects to the term "descriptive" as applied to the 
ordinary Indo-European nomenclature. When the Norwegian 
combines the stem for "father" or "mother" with that for "brother" 
to form the words farbror and morbror for the paternal and ma- 
ternal uncle, he is evidently "describing" the relationship in 
Morgan's sense of the word. But where English does not employ 
classificatory terms, as in the case of "uncle," it evidently falls 
back for the most part upon such primary stems as those for fa- 
ther or mother, which cannot by any criterion be called "descrip- 
tive," but are simply, as Rivers contends, "denotative." Morgan's 
term, then, should be restricted to systems that actually exhibit a 
strong tendency to define relations by descriptive compounds. 

So far as I am aware, no one has called attention to a basic 
logical error in Morgan's dichotomy, quite regardless of the rele- 
vant facts. "Classificatory" and "descriptive" are not comple- 
mentary concepts, but belong to different logical universes: the 
former envisages the singularity or plurality of the kinsfolk des- 
ignated; the latter considers the technique by which kinsfolk 
are defined. It is conceivable that a tribe should designate the 
paternal uncle by a descriptive phrase and apply that term to the 
whole class of father's clansmen within, say, the latter's genera- 
tion. What is more, this might even hold for so common a concept 
as that of a sibling. The Ewe call a brother "mother's-child-male." 
What is to prevent them from applying this compound as widely 
as the more usual primary stem for brother? The logical com- 
plement of "classificatory" is evidently not "descriptive" but "in- 
dividualizing"; the logical complement of "descriptive" is Rivers' 
"denotative" if the word is understood to refer to designation by 
primary stems. That this is not a matter of mere logic-chopping, 
appears from the Lango relationships, which, "though based on 
the classificatory system, include a number of descriptive terms 

' Ibid., -pp. 12, 48. 



A Islote on Relationship Terminologies 

some of which, nevertheless, are used in a classificatory way." ^ 
In this connection it seems worth while to point out another 
confusion of thought. It is often stated that classificatory systems 
are characterized by the discrimination of elder and younger sib- 
ling. Obviously, it is the more generic terms "brother" and "sister" 
that come closer to the "classificatory" standards. It so happens that 
the discrimination frequently occurs in non-classificatory ter- 
minologies. There is thus neither logical nor empirical warrant 
for the correlation asserted. After the distinction has once been 
established, the terms can of course be extended in a classifi- 
catory sense. But the distinction as such is in conflict rather than 
in harmony with classificatory ideals. 

Among the most lamentable phenomena in the recent literature 
of the subject is the tendency of British writers to speak of 
"Clan" and "Family" nomenclatures. I myself believe in a fairly 
high correlation of clan systems with a classificatory terminology 
of the Iroquois-Dakota type. However, correlation does not imply 
a hundred per cent correlation nor a cause-and-effect nexus; it 
means, on the face of it, a functional relationship in the math- 
ematician's sense. The terms here criticized are inexcusable 
because they prejudge a theoretical problem by injecting the in- 
ferred cause into the description of observed phenomena. The re- 
sult is inevitably baneful. As a matter of fact, there are clanless 
tribes with a "Clan" terminology; and to describe them as hav- 
ing "Clan" systems would not be conducive to clarity. 

As Kroeber long ago indicated,* kinship terminologies are not 
so many coherent "systems" but are each founded on a variety of 
disparate principles, all of which must be enumerated for a 
complete definition. Where the mother's sister is called "mother" 
and the sister's son (woman speaking) is a "son," these two 
features are parts of one system. But if they are linked with the 
use of separate words for "mother" by men and women, that is no 
longer part of the same organic whole. It is even virtually de- 
monstrable that particular terminologies have become less sys- 

^ J. H. Driberg, The Lango, a Nilotic Tribe of Uganda (London, 1923), p. 180. 
*A. L. Kroeber, "Classificatory Systems of Relationship," RAI, Journal, XXXIX 
(1909), 77-84. 


Kinship and Social Organization 

tematic. Thus, a term for paternal aunt implies as its logical 
correlate a separate term for brother's son (w. sp.), but in some 
tribes that term does not exist, having been lost, as we may infer 
from a comparison of cognate languages. That is to say, there 
has been a secondary departure from the systematic character 
of part of the nomenclature. The whole becomes proportionately 
harder to define in brief compass. 

If the terminologies of the world were both extensively and 
intensively better known, it would be necessary to attempt a 
wholesale classification on the basis of as many categories as 
possible. At present this is hardly feasible, and a provisional 
survey of the ground is best essayed with as simple a scheme as 
can be applied, to wit, by taking a single significant criterion. 
The historical development of the subject, from Morgan down, 
suggests the treatment of collateral relatives of the first ascending 
generation as the most suitable basis. The logical possibilities are 
the following: 

1. Uncles and aunts may be treated as parents. 

2. The paternal uncle may be classed with the father, while the 
maternal uncle is designated by a specific term; and, correspond- 
ingly, the maternal aunt may be classed with the mother, while 
the paternal aunt has a specific designation. 

3. The paternal and maternal uncles (or aunts) are alike dis- 
tinguished from the parent and from each other. 

4. The paternal and maternal uncles (or aunts) are alike dis- 
tinguished from the parent, but bear a joint uncle ( or aunt ) desig- 

The merging of uncles and aunts with parents constitutes a 
Generation terminology. If the males (or females) of the first 
ascending Generation are dichotomized on the principles ex- 
plained, the terminology may be called Bifurcate Merging: bi- 
furcate, because paternal and maternal kin are distinguished, 
merging insofar as there is a partial merging with the parents. 
Where this merging fails to obtain, so that each collateral relative 
is distinguished, the nomenclature becomes Bifurcate Collateral. 
If, finally, the collaterals are confounded with each other but 
remain separate from the direct line of descent, such emphasis on 
the latter merits the term Lineal. 


A Note on Relationship Terminologies 

Evidently, the Generation type corresponds to Morgan's mis- 
named "Malayan" or Rivers' Hawaiian system; the Bifurcate 
Merging, to the more common "classificatory" form variously called 
"Turanian- Ganowanian," "Clan," "Dakota-Iroquois"; the Lineal, 
to the common Indo-European (Morgan's Descriptive, Rivers' 
Family) system. The Bifurcate Collateral, a second "Family" 
system, has been generally ignored by theorists, though its pres- 
ence in North America has been repeatedly noted. 

The designations here employed are awkward but serve to 
bring out connections usually disregarded. Specifically, Morgan 
and Rivers stressed the genetic relationsliip of Generation and 
Bifurcate Merging terminologies: Morgan traced the develop- 
ment of the latter from the former; while Rivers reversed the 
process. The recognition of Bifurcate Collateral terminologies 
opens a new prospect, — the derivation of the Bifurcate Merging 
from the Bifurcate Collateral type. Logically, the ajffiliation is not 
one iota smaller than between the two systems compared by 
Morgan and Rivers. Empirically, and bringing in sociological 
correlations, marriage is even in clanless societies often a con- 
tract between two families, whose separateness is emphasized 
and may thus find expression in language. When, for some reason 
— say, the joint prevalence of the levirate and the sororate — 
partial merging develops, the Bifurcate Merging type would come 
into being. 


The Omaha and Crow Kinship 

In his survey of North American kinship nomencla- 
tures Lewis H. Morgan discovered that a series of Southern Siouan 
tribes, including among others the Omaha, Iowa, and Osage, as 
well as the Siouan Winnebago and a series of Algonkians, in- 
cluding among others the Illinois, Sauk-Fox, and Shawnee, did 
not classify cross-cousins according to the Iroquois terminology 
most familiar to him.-^ Instead of designating cross-cousins by a 
distinctive word, these tribes put them either into the first as- 
cending or into the first descending generation. The mother's 
brother's son was called by the same term as the mother's brother; 
indeed, this term was applied to all male descendants of the ma- 
ternal uncle through males, irrespective of generation. Cor- 
relatively, the father's sister's son was equated with the sister's 
son, which meant that a man called this cross-cousin by a distinct 
nephew term, while a woman called him her son. The mother's 

Verhandlungen des XXIV. Internationalen Amerikanisten-Kongr esses, Hamburg, 
7. bis 13. September 1930 (Hamburg, 1934), pp. 103-108. 

^ L. H. Morgan, Systems of Consanguinity and Affinity (Washington, 1871), 
pp. 179, 217. 



Omaha and Crow Kinship Terminologies 

brother's daughter was classed with the mother and logically 
called her father's sister's children as she would her own children. 

These findings were not only corroborated by later investigators 
so far as the same tribes were revisited, but a second centre of 
distribution was discovered by E. W. Gifford in California.^ It 
takes in the Miwok, Wintun, and part of the Porno and Yokuts. 
Gifford's and Spier's maps^ show that within these two widely 
separated areas the distribution is absolutely continuous. The two 
groups in question stand out sharply from all neighboring tribes by 
their peculiarities in kinship classification and are thus rightly seg- 
regated by Spier as constituting a distinct "Omaha" type. There is 
only one conceivable explanation of the distributional data — his- 
torical connection within each of the two areas. Further, we 
must note that the grouping does not coincide with linguistic 
classification: only some of the Siouans and some of the Algonkians 
help constitute the Omaha type in the East, while in California 
the Southern Pomo and the Wappo are separated from the other 
Pomo divisions. In other words, while the distinctive features of 
the type may go back to inheritance of an ancestral system so far 
as the Southern Siouans or the Central Algonkians in question 
are concerned, there must have been borrowing by the former 
from the latter, or vice versa; and similarly for the California 

On the other hand, nothing warrants us in assuming that the 
Californians borrowed their system from the East or the East- 
erners theirs from California. There is no indication that the 
Miwok and the Omaha were ever nearer to each other than they 
are now, and none of the intervening tribes show a trace of the 
features peculiar to their nomenclature. This conclusion, however, 
fails to shed any light on the resemblance. How could such an 
anomalous set of features, one in apparent contravention of 
common sense, come into being in two distinct regions? Indeed, 
the query may be widened geographically, for from outside 
America at least two clear-cut parallels are available. The Lhota 

" E. W. GifFord, Miwok Moieties, Univ. Calif. Publ. Amer. Arch, and Ethn., XII 
(1916), 172; idem, California Kinship Terminologies, ibid., XVIII (1922), 163. 

* Leslie Spier, The Distribution of Kinship Systems in North America, University 
of Washington Publications in Anthropology, I, No. 2 (1925), 69-88. 


Kinship and Social Organization 

Naga of Assam apply the word "omo" to the mother's brother and 
to the mother's brother's son; and as among the Miwok the 
mother's brother's daughter is classed with the mother's younger 
sister. Further, the father's sister's son and the sister's son are both 
designated as "orrho." * The Lango of the Upper Nile likewise 
have one word, "okeo," for the sister's and the father's sister's 
son of a man, though they do not merge the concepts of maternal 
uncle and maternal uncle's son.^ One Asiatic and one African 
parallel are thus added to our list, and we are forced to speculate 
as to the conditions that might favor the recurrence of such od- 
dities of terminology in four remote centres. 

What, then, is the element of social organization that allies 
the two groups of American tribes, the Assamese, and the Nilotics? 
A glance at Professor Spier's roster at once shows that virtually 
every one of the American tribes is patrilineal, and this likewise 
holds for the Naga and the Lango. The only exception is provided 
by the four subdivisions of the Pomo, but even for them no one 
has claimed a full-fledged maternal clan system. The most that 
can be asserted is that the Pomo have matrilineal tendencies. Be- 
sides, the proximity of tribes having the Miwok peculiarity would 
amply account for its extension to the Pomo by diffusion. 

A patrilineal system would inevitably place in the same clan or 
lineage the maternal uncle and his son; the mother, mother's sis- 
ter, and maternal uncle's daughter; the sister and the father's sis- 
ter; and as a corollary from the last equation the daughters of 
these relatives might also easily be treated as identical. The aber- 
rant classification is thus apparently nothing but an overriding 
of the generation principle in favor of the clan or lineage prin- 
ciple: kinsfolk, even though not of the same generation, are 
designated by one term if in the same unilateral group or if off- 
spring of members of the same group. 

It is possible here to apply the logic of the "conti'ol" method 
characteristic of the exact sciences. Let us eliminate the clan 
factor and see what happens. The Omaha, as shown above, rec- 
ognize an infinite series of maternal uncles, so that the mother's 
brother's son's son is likewise a maternal uncle. Let us, then, com- 

J. P. Mills, The Lhota Nagas (London, 1922), p. 93. 
■J. H. Driberg, The Lango (London, 1923), p. 177. 


Omaha and Crow Kinship Terminologies 

pare with this the designation for the mother's brother's daughters 
son. We are keeping the sex and generation of the relative the 
same, only varying his clan affiliation, since paternal descent would 
place him in a different clan from his mother's. The Omaha re- 
spond to the change of stimulus with the precision of an automa- 
ton. The mother's brother's daughter's son is not a mother's brother. 
His status is determined by the simple fact that his mother is 
considered the speaker's "mother," hence he becomes a brother. 

Mother's brother's daughter = mother 

(Mother's brother's daughter )'s son = mother's son 

Whence, by substitution: mother's son = brother 

But while the Omaha-Miwok type thus emerges as functionally 
connected with paternal descent, paternal descent by itself can- 
not be regarded as an adequate determinant. In America there 
are patrilineal tribes, like the Cree and Ojibwa, who do not share 
the Omaha peculiarities and, as a matter of fact, are aligned with 
matrilineal tribes like the Seneca and the Tsimshian.^ We must 
accordingly seek an additional factor that might effect the ob- 
served phenomenon. 

The supplementary determinant required appears both among 
the Omaha and the Miwok. The latter allowed a man to marry 
his wife's brother's daughter before or after his first wife's death. 

"In some cases, if she were too young for him to marry, she 
was held for him until she had reached the marriageable age, 
when she was handed over to him." ^ 

Similarly, an Omaha may marry his wife's brother's daughter 
(as well as her father's sister or her sister, elder or younger).^ 

Gifford has pointed out how an institution of this sort would 
affect terminology ® and a variant of his diagram will serve to il- 
lustrate the point. 

Here, the small letters indicating females and capitals males, 
A appears as the husband of both b and her niece e. Let us then 
view the situation from the angle of the children he begets with 
the younger wife. For g or H, F = a mother's brother (hterally). 

" Spier, op. cit., p. 78. 

'' Gifford, Miwok Moieties, p. 186. 

'J. O. Dorsey, "Omaha Sociology," BAE, 3d Ann. Kept. (1884), pp. 257, 261. 

* Gifford, California Kinship Terminologies, p. 248. 


Kinship and Social Organization 

But since A is married to b likewise, b as well as e figures as their 
mother, hence C no less than F is their "mother's brother." But 
F is literally C's son. Hence, F is at the same time g's and H's 
maternal uncle and their maternal uncle's son; which was to be 
proved. Further, since e has become A's wife, the children i and K, 
whom he has begotten with b, become e's children. Shifting, 
then, the point of origin to F, his sister's children in this extended 
sense are identical with the actual children of his father's sister. 

A=b C=d 

K A = e 

g H 

Mr. Driberg has shown that the Omaha-Miwok feature shared 
by the Lango has a diflFerent origin, so that we apparently meet 
here one of those rare instances of demonstrable convergence. Like 
many African Negroes, the Lango practise filial succession as to 
widows. Accordingly, a man F will anticipate his status as d's 
husband, equate himself with his father, hence look upon b as 
his sister, and upon b's children as his sister's children. Of course, 
the widows inherited do not include the actual mother, but a 
father's wife addressed as mother.^'* 

The forms of marriage deemed orthodox by the Lhota are very 
interesting. A man may take his brother's widow without bride 
price, is expected to marry into his mother's clan and is fined "if, 
having taken one wife from his mother's clan, he takes a new one 
from another clan." He may marry his mother's brother's daughter 
but not the other cross-cousin; we do not learn whether his wife's 
brother's daughter is a possible mate. However, "He may also 
marry his father's widow provided tliat she is not his own mother, 

" Driberg, op. cit., p. 186. 


Omaha and Crow Kinship Terminologies 

but such marriages, though pretty common, are viewed with a 
certain amount o£ disfavor." According to a footnote by Mr. I. H. 
Hutton, this quahfication does not hold for the Sema Naga.^^ 

In each of our four areas, then, there is not merely paternal 
descent, but a supplementary factor, viz., marriage with mem- 
bers outside of one's generation, and it is the leveling influence 
of this determinant that produces the Omaha-Miwok features. 
When, however, we consider these forms of marriage, it is clear 
that they are themselves functions of paternal descent. It is in 
patrilineal tribes that a son inherits his father's widows; it is 
patrilineal tribes that would substitute for a woman her brother's 
daughter (Omaha, Miwok) or her father's sister (Omaha). What 
we are here confronting is merely a somewhat unexpected but by 
no means especially rare extension of a familiar principle. Tylor 
has familiarized us with the idea of primitive marriage as a com- 
pact of groups rather than of individuals. Individuals of the same 
group are reckoned equivalent: sisters from one household are 
exchanged for sisters of another, a deceased wife is superseded by 
her sister, a deceased husband by his brother without further 
payments, etc. We can easily understand that in small communities 
such substitutions might be difficult if one rigorously adhered to 
the principle of equivalence within the generation. We can also 
understand that the economic value of women might stamp them 
as the most valuable portion of a man's estate and that the heir 
is often bound to be of a lower generation. As soon as the notion 
is waived that social equivalence holds only within the genera- 
tion the Omaha-Miwok Lango customs spring into being. What 
relatives shall be deemed equivalent will depend largely on 
whether there is a maternal or a paternal bias. 

There is nothing in the nature of things that would inhibit 
such extensions of the ideas underlying the levirate and the so- 
rorate to matrilineal peoples. We might accordingly expect to 
discover corresponding phenomena among matrilineal tribes, and 
this is justified. Morgan found that a comparable cross-cousin 
nomenclature occurred among the Crow, Hidatsa, Choctaw, 
Creek, Cherokee, Pawnee, and Laguna; in other words, in the 
Northern and Southern Plains, the Southeast, and the Pueblo 

" Mills, op. cit., pp. 95, 155. 


Kinship and Social Organization 

areas. ^~ These tribes all possess one or all of the following traits; 
they call the father's sister's son by the same word as the father; 
the father's sister's daughter and her female descendants through 
females ad infinitum are all classed with the paternal aunt; the 
mother's brother's son is a son, and the mother's brother's daughter 
a daughter ( more frequently for a male than a female speaker ) ; 
the mother's mother's brother is equated with the maternal uncle. 
Rivers pointed out that Codrington had recorded a similar scheme 
in the Banks Islands ( Melanesia ) ; ^^ and our present knowledge 
of relevant American data has been succinctly summarized by 
Spier.^^ From Spier's statement we learn that the features in ques- 
tion also occur on the Northwest Coast, among the Tlingit and 
Haida; and in California, among the Southern Pomo and Wappo. 

Not one of the tribes listed is patrilineal; the majority have 
definitely maternal clan organizations; the Pawnee, Pomo, and 
Wappo are credited with a definite matrilineal bias. The Banks Is- 
landers are organized into maternal clans. It is obvious that with 
maternal descent my father's sister's son and daughter are bound 
to be in my father's and my father's sister's clan; also that the 
female descendants of my paternal aunt through females will al- 
ways be of her clan. It is equally clear that if descent were pater- 
nal these alignments would no longer hold. With two exogamous 
clans, e,g,, if my father is of clan I, his sister is also I, but her 
husband necessarily belongs to II, and the father's sister's chil- 
dren are II. The agreement of the Crow cross-cousin scheme with 
matrilineal reckoning is thus perfect. 

Again, we may take a test case. What happens if we eliminate 
the clan factor by taking the father's sister's sons daughter in- 
stead of the father's sister's daughter's daughter? The Crow are 
no less logical than the Omaha: 

father's sister's son = father 

(father's sister's son)'s daughter = father's daughter 

father's daughter = sister 

The daughter of the male cross-cousin defined above is thus not 
a paternal aunt, but a sister. In short, the Crow peculiarity is a 
function of matrilineal descent. 

' Morgan, op. cit., pp. 188, 191, 197, 262. 

W. H. R. Rivers, Kinship and Social Organisation (London, 1914), pp. 28, 53. 
' Spier, op. cit., p. 74. 


Omaha and Crow Kinship Terminologies 

But here we are met by a diflBculty exactly like that in the 
Omaha case. Not all matrilineal tribes have systems of the Crow 
type. Most of the Iroquois and the Tsimshian lack the traits char- 
acteristic of this type. What is the differential condition that makes 
for the Crow peculiarities? The Omaha-Miwok cases suggest 
looking for a special form of marriage that might produce the ob- 
served equations, and Rivers offers a solution in consonance with 
the Melanesian data. In the Banks Islands a man inherits his ma- 
ternal uncle's wife. The result is plain. 

A=b C=d=E 

E f G h 

From the point of view of G, the son of E's maternal uncle, 
E is his (G's) father's sister's son. Since this cross-cousin inherits 
d, who is G's own mother, E becomes G's father and is thus 
simultaneously his father and his father's sister's son. But inas- 
much as E is now a father to G, f — being E's sister — is promoted 
to the status of father's sister to G 

In pure logic this explanation is unexceptionable. Empirically 
it is not adequate, for we cannot as yet demonstrate the inher- 
itance of a man's widow by his uterine nephew except in Melanesia 
and on the Northwest Coast of America. The Crow and Hidatsa 
might conceivably be added because of a feature of their systems 
unique in North America: like the Melanesians of Leper's Is- 
land ^^ they lack a term for mother's brother, classing him as an 
elder brother — a classification once more intelligible on the theory 
that clan considerations may outweigh the principle of genera- 
tions. Though we do not actually know that Crow and Hidatsa 
nephews inlierited their maternal uncle's widows, such inher- 
itance, as Dr. Lesser has pointed out ^^ would thus be a simple 
application of the levirate, which in the usual form is practised 

'^ A. B. Deacon, "The Regulation of Marriage in Ambrym," KM, Journal, LVII 
(1927), 327. 

" Alexander Lesser, "Kinship Origins in the Light of Some Distributions," Amer- 
ican Anthropologist, XXXI (1929), 720. 


Kinship and Social Organization 

by both tribes. However, for the three remaining areas — the 
Southern Plains, the Southeast, and the Southwest — I have not 
been able to ascertain marriage with the uncle's wife or other 
unions with members of a higher generation. This may be due 
merely to defective observation, but that cannot be taken for 
granted. Hence, we must again cast about for alternative solu- 

At least one emerges from recently recorded Navajo data. The 
special peculiarity registered for the Crow, Hidatsa, and Hopi of 
merging the elder brother, maternal uncle, and mother's mother's 
brother terms ^^ would be explicable from the Navajo custom by 
which a man marries his wife's daughter by a previous mar- 
riage. ^^ In such a case, the child, g or H, borne by the daughter, 
f, has for a second "mother" his maternal grandmother, b, since 
she is his father's, A's, wife. Since the mother's mother = the 
mother, the mother's mother's brother = the mother's brother. 
Further, since 

C = b's son and 
also C = f's son, because she marries his father, 

C = H's brother 
Actually, C = H's mother's (mother's son) 

C = H's mother's brother 
Whence, mother's brother = brother 

But from tliis the most distinctive features of the Crow type 
directly follow: 

mother's brother's son = brother's son = (man speaking) son 
Whence, by correlation, father's sister's son = father 
father's sister's daughter = father's sister 

In short, the Crow type is deducible from Navajo marriage 
arrangements as well as from nepotic widow-inheritance. Un- 
fortunately we are not informed as to the distribution of the cus- 
tom permitting simultaneous marriage with a woman and her 
daughter. The Navajo themselves, it must be admitted, do not 
explicitly carry out the logical implications of the arrangement. 

" R. H. Lowie, Notes on the Social Organization and Customs of the Mandan, 
Hidatsa, and Crow Indians, AMNH, Anthropological Papers, XXI (1917), 59. 

" G. A. Reichard, Social Life of the Navap Indians, Columbia University, Con- 
tributions to Anthropology, VII (1928), 59, 62 ff. 


Omaha and Crow Kinship Terminologies 

However, Dr. Reichard's genealogical data demonstrate a rather 
frequent shifting of generation status, individuals being some- 
times raised or lowered two degrees. Thus, one of her subjects 
calls his father's mother's mother's brother's son his elder brother 
— ostensibly because the addressee and the speaker have fathers 
of the same clan who are accordingly rated as brothers/^ But with 
maternal descent every man is in the same clan as his mother's 
mother's brother, hence the stressing of clan alignment would 
favor identification of this relative with the elder brother, which 
actually occurs in explicit form among the Hopi, next-door neigh- 
bors of the Navajo. It seems far from improbable that the fre- 
quent departures of Navajo nomenclature from generation lines 
are correlated with their highly distinctive form of marriage. 

Perhaps there are still other matrimonial regulations that might 
produce the Omaha or the Crow type of nomenclature. At all 
events, the general problem of these atypical systems may be 
formulated as follows. When a tribe practises the levirate and so- 
rorate or is organized into clans and recognizes the matrimonial 
equivalence of generation-fellows within a clan, the kinship ter- 
minology is likely to follow the Dakota-Iroquois type. As Dr. 
Lesser and Mr. Deacon have suggested independently, when 
equivalence is extended to members of higher or lower genera- 
tions the Omaha or the Crow type evolves. ^^ But which of the 
two? That will depend, I suggest, on the rule of descent, for the 
relatives who may be substituted, or who are added as fellow- 
spouses, are selected on that principle. We do not find a Navajo 
woman o£Fering a father's sister or a brother's daughter to her hus- 
band as a supplementary wife; and correspondingly it is not the 
Omaha wife that makes of her daughter a fellow-spouse. 

In some earlier publications I inclined to explain kinship ter- 
minology as far as possible from the clan only, to the virtual ex- 
clusion of specific marriage rules. I thus took issue with Rivers, 
and was opposed by Sapir, who stressed the levirate and sororate, 
and Gifford, who was impressed with the marriage of a man with 
the daughter of a wife's brother. Quite recently, Dr. Durlach has 

" Ibid., p. 83. 

^Lesser, op. cit., p. 711. Deacon {op. cit., p. 328) regards inheritance of a 
maternal uncle's widow as an extension of the levirate. 


Kinship and Social Organization 

again taken me to task for assuming "too general" a point of view: 
"the reasons why certain terms should be extended beyond others, 
why some should override generations, and others not, must after 
all be sought in more specific causes, such as marriage regula- 
tions, social functions, or language." ^^ While I do not understand 
the last of these suggested factors, I have long been in sympathy 
with the remainder of the statement.^^ What I insist upon is 
that the more specific matrimonial arrangements are themselves 
a function of the rule of descent. For example, in her recent 
publication Dr. Durlach has, so far as I know, been the first to 
register certain intergeneration forms of marriage. The Tlingit 
frequently married their elder brother's daughters, and a union 
with one's father's sister was preeminently proper.^^ These ar- 
rangements characteristically occur in a matrilineal society. 

^ T. M. Durlach, The Relationship Sijstems of the Tlingit, Haida and Tsim- 
shian, American Ethnological Society, Pviblications, XI (1928), 13. 
^ R. H. Lowie, Primitive Societij (New York, 1920), p. 37. 
^ Durlach, op. cit., pp. 64 f. 



The Family as a Social Unit 


parents and their children, whether actually living together or 
not" (Murray's dictionary). The concept may be enlarged to 
embrace "those who are nearly connected by blood or affinity," 
but such expansion makes for greater vagueness. Adhering to the 
narrower definition, let us ask whether human society must a 
priori be constituted of family units. The answer is negative. There 
are sexually reproducing species without a semblance of family 
life, hence the segregation of husband, wife, and child into a 
distinct group remains to be empirically demonstrated. As a mat- 
ter of fact, the existence of such a unit in early man has been 
categorically denied by many writers. In the beginning, we are 
told, was promiscuity — sexual license unchecked by any restraint. 
The earliest inhibitions prevented interbreeding of parent and 
child; they were followed by interdicts against the union of sib- 
lings ( brother and sister ) ; and so by a series of reformatory move- 
ments humanity finally attained the giddy heights of Victorian 
monogamy, at least in theory. 

Unfortunately, we can know nothing directly about the sex life 

Papers of the Michigan Academy of Science, Arts, and Letters, XVIII ( 1933 ) , 


Kinship and Social Organization 

of man's immediate precursor, and a comparison of primate be- 
havior, though definitely ruhng out certain assumptions, offers 
a minimum of positive fact for the reconstruction of ancestral 
habits. A zoologist, Mr. Gerrit S. Miller, Jr., has recently brought 
together what is known. He has proved that, contrary to a wide- 
spread misconception, virtually all the primates observed lack 
a rutting season. Accordingly, it is in the highest degree improb- 
able that man's immediate forerunner mated seasonally. Like his 
fellow-primates, he was presumably ready to indulge in amours 
whenever an occasion arose. Furthermore, it seems that recent 
human aberrations have their counterparts among primates, and 
their potentiality may thus well be a heritage from a fairly remote 

From available information, however, we can gather nothing 
to test the theory of early human promiscuity. Indeed, the field 
observations on the nearest anthropoids, chimpanzees and gorillas, 
are indecisive and at times contradictory as to the traits of the 
same species. Reichenow, for example, credits the gorilla with 
monogamous habits, while Akeley cautiously suggests the pos- 
sibility of polygamy. "The truth is," he wisely adds, "that people 
know little about the habits of the gorilla." ^ 

Yerkes, with exemplary restraint, makes the following state- 

Our tentative inference is that both monogamy and polygamy exist 
in one or another or all of the anthropoid types and that in all prob- 
ability both relationships are discoverable in each of the manlike 
apes. With many misgivings we propose as order of increasing prob- 
ability of monogamic relation: gibbon and siamang, gorilla, orang- 
outan, chimpanzee. Much more systematic, thorough, and critical 
investigation than has heretofore been conducted will be essential 
to discover the truth. Indicated as points of contrast among the three 
types of great ape are temporary monogamous or polygamous rela- 
tions in the orang-outan, relatively permanent monogamous and 

^ G. S. Miller, "Some Elements of Sexual Behavior in Primates and Their Possible 
Influence on the Beginnincrs of Human Social Development," Journal of Mammalogy, 
IX (1928), 273-292; idem, "The Primate Basis of Human Sexual Behavior," 
Quarterly Review of Biology, VI (1931), 379-410. After writing this paper, I find 
that Miller's inferences are challenged by Dr. Solly Zuckerman in Social Life of 
Monkeys and Apes (New York, 1932). The matter is one for zoologists to decide. 

^Carl E. Akeley, In Brightest Africa (Garden City, 1925), p. 247. 


The Family as a Social Unit 

possibly also polygamous relations in the chimpanzee, and in the 
gorilla a patriarchal family, with polygamy presumably in the moun- 
tain species and monogamy, possibly, in the lowland species. ^ 

If we know nothing more positive about existing species, any 
dogmatic conclusion as to the behavior of a hypothetical, ex- 
tinct ancestral type seems rash indeed. 

On one point, however, we can be certain. Whatever may have 
been the mating habits of this or that precursor of Homo sapiens, 
no believer in evolution can deny a stage of promiscuity some- 
where along the line, that is, of promiscuity in the technical sense 
of socially unrestrained lust. Anthropologically, there is no "index 
of promiscuity," calculated by dividing the number of actual 
mates, regardless of kinship, by the number of physically possible 
ones. From this angle, it is a question of "all or nothing." Is carnal 
desire checked in some of its manifestations by the disapproval 
of the group? If it is, there is no promiscuity; otherwise, there is. 
Take the case of a male gorilla which Akeley found with three 
females. The point is not whether the male cohabited with all 
three females. It is rather this: Assuming two of them to be his 
daughters, would the attitude of other gorillas be one of indif- 
ference or not? The situation is not inconceivable even on the 
human level. A widespread tale of Great Basin and Western 
Plains Indians revolves about this very theme. The trickster by his 
wiles gains access to his own daughters. In the story, however, 
such behavior arouses intense moral condemnation. Now, I, for 
one, fail to find evidence of such a social consciousness in either 
Koehler's * or other data from the infrahuman plane. If this in- 
terpretation holds, the anthropoids are promiscuous. On the other 
hand, no known group of Homo sapiens is indifferent to the sex 
behavior of its constituent members. Wherever evidence is ade- 
quate, matings are judged — outlawed, reprobated, condoned, ac- 
cepted, or definitely sanctioned. The definitely sanctioned forms 
of mating may be termed "marriage," and from them evolves the 
family. Nowhere are fornication and marriage submerged in an 
undifferentiated category of animal-like "copulation." 

^ Robert M. Yerkes and Ada W. Yerkes, The Great Apes, a Study of Anthropoid 
Life ( New Haven and London, 1929 ) , pp. 542 f. 

* Wolfgang Kohler, The Mentality of Apes (New York, 1925). 


Kinship and Social Organization 

A chasm thus yawns between Homo sapiens and the chimpan- 
zee or gorilla. At what stage of evolution, then, was the leap 
taken from unjudged to judged sex behavior? I do not know. 
I venture a guess that Neanderthal man showed some discrimina- 
tion. I so conjecture because he demonstrably had a social tradi- 
tion as to craftsmanship, and it thus seems probable to me that 
he had likewise evolved norms of social conduct. I refuse even to 
guess whether Heidelberg man, Eoanthropus, Peking man, and 
Pithecanthropus displayed equal fastidiousness. I am content to 
believe that, somewhere between the more remote anthropoid 
ancestor and the more immediate hominid ancestor whose de- 
scendants constitute geologically recent humanity, there was a 
stage of uncontrolled sexual license. 

I am not sure whether I agree or disagree with Mr. Miller as 
to the distance of this stage. He oflFers the argument that living 
samples of men are specialized survivors and that many races 
have become extinct. Hence, he infers, "the search among these 
specialized existing peoples for a race or tribe living under social 
conditions that represent anything closely resembling an un- 
modified reflection of man's primitive mentality can have little 
chance of success." ^ Here everything hinges on the meaning of 
the terms "man," "closely," "primitive mentality." I not merely 
admit but contend that Andamanese, Fuegians, Australians, and 
Chukchi tell us nothing definite about the mentality of Piltdown 
or Peking man. I emphatically insist that no one primitive group 
represents the first hominid's mentality in unmodified form. But 
if such highly specialized groups as Andamanese, Australians, and 
others, without exception exercise social control of sex life, then 
such control does not date back to yesterday nor, say, to 4000 
B.C., but, presumably, to a period embracing the earliest samples 
of Homo sapiens, even though some of the races of this species are 
irrecoverably removed from direct observation. 

Time does not pemiit detailed consideration of more than one 
recent human society. I shall select the Australians, whose ana- 
tomical inferiority and crudeness in the arts of life have made 
them a favorite starting point for speculative liistorians on the 
origins of the family, religion, and what not. Moreover, they 

^ Miller, "The Primate Basis of Human Sexual Behavior," p. 400. 


The Family as a Social Unit 

have been credited with a form of sex hfe that might be viewed 
as intermediate between promiscuity and obhgatory monogamy, 
viz., "group marriage." This institution has been defined as the 
non-preferential mating of a group of men with a group of women. 
It would not represent promiscuity, inasmuch as Australians would 
never tolerate unions of brothers and sisters. But anyone who 
favors the theory of promiscuity in Aurignacian or Mousterian 
times would naturally regard that mixture of polyandry and po- 
lygyny involved in group marriage as a step toward increasing 
control of mating. On the other hand, so long as a whole group 
of men mated indiscriminately with a group of women, the family 
would remain non-existent as a social unit. 

In the interests of concreteness I shall base my statement on 
what Professor Radcliffe-Brown describes as the clearest available 
account of Australian conditions, Warner's report on the Murngin 
living west of the Gulf of Carpentaria.*' This picture I shall eke 
out with supplementary data on the Australians and shall then 
proceed to cull relevant data from the literature on other groups. 
The questions asked will include the following: Is there a form of 
marriage as distinguished from cohabitation? If so, what are the 
social relations of husband and wife? Of siblings? Of parent and 

To begin with the Australians, no Murngin is free to mate with 
whom he pleases, and in marriage he is always expected to obtain 
the daughter of his maternal uncle. Failing such a one, a substitute 
of equivalent kinship status would be sought, for example, the 
daughter of a mother's male cousin. Potential spouses may be be- 
trothed prenatally. To be sure, a man wants the maximum number 
of wives safely procurable, but they are never chosen at random. 
In order to make social intercourse possible for them at all, Aus- 
tralians always range individuals into kinship classes. So, even 
when the Murngin raid a hostile camp the kidnaped women are 
allotted to men standing to them in the socially approved relation- 
ship. Similarly, adultery is almost always with a cousin of the 
prescribed category. By a natural extension of these ideas, which 
rest on the social equivalence of siblings of the same sex, a brother 

" William Lloyd Warner, "Morphology and Functions of the Australian Murngin 
Type of Kinship," American Anthropologist, XXXII (1930), 207-256. 


Kinship and Social Organization 

inherits his elder brother's widow, and often the several wives of 
a polygynous husband are sisters or quasi-sisters. 

Unquestionably there are "wrong" marriages among the Murn- 
gin, as among ourselves. Yet within certain degrees prohibitions 
are absolute and, apparently, never flouted. In other cases strong 
disapproval is voiced: a man who would carry on an intrigue with 
his "sister's daughter" — actually perhaps his third cousin's daugh- 
ter — would be compared to a dog, and the woman would be 
liable to a severe drubbing. 

Moreover, within the range of licensed unions a distinct ideal 
may be noted. A husband may have several wives, but he ought not 
to seek amours with other women; and a wife is normally expected 
to content herself with a single mate, her husband. The social 
relations of spouses, furthermore, assume definite rights and duties. 
A wife gathers wild fruits and small game; the man supplies fish, 
turtle, porpoise, dugong. Sentimentally, common devotion to the 
children constitutes a bond; and even apart from that factor in- 
dications are not lacking of an attachment reminiscent of romantic 

In all this there is not the faintest suggestion of either promis- 
cuity or group marriage. The parental relationship is extended so 
that a woman's sister may help her suckle two babies; and, in gen- 
eral, a child looks to a maternal aunt for food and care. This, how- 
ever, develops quite naturally from the practice of sororal po- 
lygyny. But, though the principle of sibling equivalence holds, 
the immediate family group is distinguished. Thus, a childless 
husband observes food taboos, which are lifted with paternity, 
"but the child must be his own, not that of a brother." (The term 
"own" in this context will be discussed later. ) It is the father who 
determines the type of initiation for his son, passes on the right 
to certain dances, and teaches the ceremonial routine. In short, 
a man takes a differential interest in "his" children. 

The Murngin thus recognize a family unit, but that does not 
mean that it is our family pattern. A contrast at once appears with 
regard to siblings. Brother and sister can never be on terms of 
easy familiarity. A brother never sleeps in the same camp with her, 
and neither may address the other. Associated with such taboos 


The Family as a Social Unit 

we find the attitude of mutual helpfulness that to us seems alto- 
gether intelligible. A brother will give presents to his sister for 
her son and husband. Two brothers cooperate in economic pur- 
suits and have a sense of joint ownership of property. This natu- 
rally in a measure embraces wives, but with such qualifications as 
to exclude unchecked communism even between true brothers. 
No younger brother appropriates a sister-in-law without permis- 
sion. The elder brother preemptively claims his maternal uncle's 
daughters. When he has thus acquired two wives, the younger man 
has a strong moral claim on the next oldest sister of the household, 
and her father may urge the husband to waive his legal preroga- 
tive. Even here there is thus definite customary law, not license. 
But a brother's attitude cannot be the same as ours in a society 
which makes him look to his older brother as the provider of a 
mate, either after or during his lifetime. 

The family picture would be further modified by the taboo, uni- 
versal in Australia, forbidding all social intercourse between a man 
and his mother-in-law. Yet, notwithstanding the lavish use of such 
kinship terms as "father," "brother," etc., to embrace fairly remote 
kinsfolk, the immediate family group is clearly separated from the 
rest of the community. A prospective husband tries, first of all, 
to marry his "own" mother's "own" brother's "own" daughter; 
and the uncle provokes resentment if he marries oflF a daughter 
to a remote nephew. 

We have seen that a married man's social status depends on 
his having "own" children. This distinction between near and 
remote kin of the same category holds throughout. Remote "broth- 
ers" ambush and slay one another, or at least suspect one another 
as potential adulterers; but between true brothers there is im- 
plicit trust and unfailing devotion. So in periods of ceremonial 
license distant, not "own," brothers participate in the temporary 
exchange of wives. Thus, at every step we stumble on clear-cut 
evidence for the aboriginal feeling that relationship to the next of 
kin is a thing sui generis. The resulting family is a bilateral unit 
since, from the child's angle, relations are maintained with both 
parental sides. 

The condition described by Warner is not unique, but typical 


Kinship and Social Organization 

for the island continent. Malinowski's synthetic review ^ o£ the 
earher hterature and Radchffe-Brown's more recent summary ^ 
leave no doubt on that point. Throughout Australia the nearest 
equivalent of our political unit, the state, is a localized "paternal 
lineage" or "horde" owning and exploiting in common a definite 
territory. Such a group embraces as a permanent core a number 
of brothers with their sons, sons' sons, and so forth. The women of 
the group normally come from another similarly constituted horde. 
Of the children the boys remain, acquiring from early childhood 
that intimate economic knowledge of the hereditary land which 
is a prerequisite to survival. The girls marry outside their horde, 
so that female children are only temporary constituents of the 
group into which they are born. Within this clearly defined horde, 
however, the aborigines recognize a lesser social unit, to wit, the 
individual family of parents and children. "The important function 
of the family," says Radclilfe-Brown, "is that it provides for the 
feeding and bringing up of the children. It is based on the co- 
operation of man and wife, the former providing the flesh food and 
the latter the vegetable food, so that quite apart from the question 
of children a man without a wife is in an unsatisfactory position 
since he has no one to supply him regularly with vegetable food, 
to provide his firewood, and so on. This economic aspect of the 
family is a most important one and it is partly this that explains 
Australian polygyny. I believe that in the minds of the natives 
themselves this aspect of marriage, i.e., its relation to subsistence, 
is of greatly more importance than the fact that man and wife 
are sexual partners. . . . sexual relations between a man and a 
woman do not constitute marriage in Australia any more than 
they do in our own society." 

I believe that the picture our foremost authorities give of Aus- 
tralian conditions may be generalized for recent races of man. 
Twelve years ago I wrote: "The bilateral family is . . . an abso- 
lutely universal unit of human society." ^ These are strong words, 
but I still regard them as essentially correct. In only one area of 

' Bronislaw Malinowski, The Family among the Australian Aborigines, a Soci- 
ological Study (London, 1913). 

^ A. R. Radcliffe-Brown, The Social Organization of Australian Tribes, Oceania 
Monographs, No. 1 (Melbourne, 1931), esp. pp. 4, 6, 11 ff., 103, 107. 

° R. H. Lowie, Primitive Society (New York, 1920), p. 78. 


The Family as a Social Unit 

the world am I able to detect phenomena tending to qualify this 
view. In parts of Oceania, where adoption plays an extraordinary 
role, children are reported to divide their time more or less evenly 
between two homes, thus participating simultaneously in two 
family groups. I have recently taken cognizance of such facts, 
writing: "In this extreme form the custom [of adoption] inevitably 
modifies the principle of the universality of the individual fam- 
ily." ^° Let us note in passing that the exceptions occur in highly 
sophisticated horticultural societies which cannot possibly be 
regarded as illustrating primeval usage; and that the exceptions 
rest on a custom which by definition is derivative. 

The general prominence of the family cannot, of course, be 
demonstrated without passing in review one primitive society after 
another, which space does not permit. I should like, however, to 
point out one rather significant North American phenomenon. 
If we examine the kinship systems of the rudest peoples in North 
America, the purely hunting tribes devoid of complex political, 
social and ceremonial organization, we find that almost uniformly 
they distinguish in speech the immediate members of the family 
from the more remote kin. That is to say, while the Australians 
recognize the distinction in behavior, the simpler North American 
aborigines go so far as to express the sense of the difference in 
their vocabularies : a father is not only treated differently from an 
uncle, but is designated by a separate term; similarly, a brother 
is not included in the same term as a cousin; and so forth. The fact 
that the majority of non-horticultural tribes from the Arctic to 
northern California and Nevada fail to merge these relatives 
strongly suggests that the family unit is clearly recognized pre- 
cisely on the lowest cultural level north of Mexico. It appears as 
though the family enjoyed undisputed ascendancy at a very 
early period, its significance being subsequently modified, though 
never abrogated, by other forms of organization. Thus, in Australia 
the partial equivalence of siblings of the same sex readily qualifies 
the character of the individual family, though its persistence is 
now demonstrated beyond cavil. 

Terms for social units, such as "family," have misleading sug- 

" Idem, "Adoption, Primitive," in Encyclopaedia of the Social Sciences, I ( 1930), 


Kinship and Social Organization 

gestiveness; therefore I shall try to indicate the empirical range 
of the data properly coming under this head. Let me first explain 
that the biological family is not necessarily identical with its 
social equivalent. A clever writer has recently credited me with 
a belief in the social omnipresence of the biological family. She 
contrasts with this the saner view of Radcliffe-Brown, who, while 
taking the biological group as the chief point of reference in a 
treatment of social organization, "gives due weight to more 
complex developments characteristic of many primitive soci- 
eties." ^^ Actually, there is no conflict; what is particularly im- 
portant, both Radcliffe-Brown and I emphatically warn against 
attaching too much weight to the biological aspect of the unit. 
"Bilateral" and "biological" are not synonymous terms. When 
an Australian speaks of his "own" father, he does not necessarily 
mean his begetter at all, but the adult male whom he preeminently 
associated from infancy with a certain emotional behavior, eco- 
nomic activities on behalf of the household, and so forth. Else- 
where I have pointed out, on Rivers' authority, that among the 
Toda of southern India polyandry often makes the deteraiination 
of paternity very difficult. But the natives do not care at all about 
biological paternity: that husband who performs a certain rite 
during his wife's pregnancy becomes legal father of all children 
borne by the woman until another husband goes through the same 
ceremony. "Biological paternity is completely disregarded, for 
a man long dead is considered the father of a child provided no 
other man has performed the essential rite." ^^ So, in some South 
African tribes a man claims as his own legal issue the offspring of 
a duly purchased wife, even if she has for years been living in 
adulterous union with a lover. What counts, then, is not the bio- 
logical but the legal kinship. The omnipresence of the bilateral 
family, then, means this: Virtually everywhere a male, who is 
not necessarily the procreator, and a female, who is not neces- 
sarily the bearer, maintain preferential relations to a given child 
or number of children, thus forming a distinct unit within any 
major social group. 

The fact that substitutions for biological parental relations are 

"Margaret Mead, "Family, Primitive," ibid., VI (1931), 65-67. 
^ Lowie, Primitive Society, p. 48. 


The Family as a Social Unit 

possible and even relatively frequent is precisely one of the most 
outstanding revelations which ethnography has to offer to her 
sister science, psychology. For it sweeps away once and for all 
the assumption of a paternal instinct. In its place we must recog- 
nize a much vaguer tendency of adult males to form an attach- 
ment to infants of their species. 

Toda and Bantu indifference to the identity of the procreator 
suflBces to mark off their conception of the family from that tradi- 
tional in Western civilization. To these natives the insistence on 
recognizing as one's children only those duly begotten by oneself 
must appear as ludicrously irrelevant physiological pedantry. Ap- 
praisal of the children's status may rest on quite different con- 
siderations. Among Northwest Calif ornian Indians the equivalent 
of the Occidental bastard is the boy whose father failed to pay the 
customary bride-price, for with that blot on his escutcheon he is 
never permitted to enter the men's club house. 

Socially, however, the family pattern is only moderately al- 
tered by the lack of interest in physiological bonds. For, in the 
examples cited, one male simply supersedes another as the em- 
bodiment of the paternal principle. In other words, a social tie of 
our own parent-child relationship category remains. That cate- 
gory may be more definitely affected by a maternal clan organiza- 
tion. Where such an institution occurs, the bond with the father 
and his kin is still recognized, but all children are, for certain 
purposes, reckoned as of kin only with the mother and, specifi- 
cally, bear the name of her clan, not their father's. In this way 
may be set up a series of sentiments, of legal rights and duties, that 
come to compete with the parental ties and even enter into open 
conflict with them. By so doing they also inevitably clash with the 
family as an autonomous social unit. This appears most clearly 
where the avunculate holds sway. There the maternal uncle usurps, 
according to our notions, many paternal functions, and, cor- 
relatively, it is his uterine nephews and nieces that often stand 
to him in a relationship we regard as filial. Thus, he, and not 
the father, may dispose of a girl's hand; he, and not the father, 
will give certain kinds of instruction to boys; and, though in 
some patrilineal African tribes, a man's son inherits his fathers 
wives, barring only his own mother, certain matrilineal American 


Kinship and Social Organization 

and Melanesia!! groups permit a nephew to marry the widow of 
his mother's brother. To take a concrete case, a Dobu in Melanesia 
cannot bequeath his name, land, status, or fruit trees to his son; 
all of them are automatically inherited by a sister's son. A man 
may i!ideed teach his son what he knows of magical formulae; but 
to his uterine nephew he must convey such knowledge. ^^ 

Nevertheless, the sociological father is not abolished by avuncu- 
lar customs. In the very region from which my last example is 
taken Professor Malinowski has demonstrated the depth of at- 
tachment linking father and son. The lurid and tragic conflict 
between paternal sentiment and avuncular duty has never been 
more vividly set forth. ^^ 

Another condition modifying the pattern of family life may be 
generalized under the head of "sex dichotomy," which manifests 
itself in many ways. Among the Australian Murngin we found 
the rather widespread custom of brother-sister avoidance, which 
at once precludes one of the most typical forms of family intimacy 
in our civilization. But we also saw that such usage does not 
snap the bond which links siblings together: brother and sister 
may not chat together, but they do aid each other, and the 
brother is keenly sensible of certain duties toward liis sisters. 
Another type of dichotomy separates husband and wife. In many 
communities, for example, in South America and Oceania, spouses 
never eat together — an arrangement almost inconceivable to us. 
Yet the Banks Isla!iders of Mela!iesia go further. Among them 
virtually every adult male has bought his way into the men's 
club house, which is strictly tabooed to women, while the men not 
only lounge and work, but eat and sleep there, paying inteiinit- 
tent visits to their wives. Notwithsta!iding this institution, the 
family still holds together, so far as a husband exercises defi!!ite 
rights over his wife a!!d is bound to her and the childre!i by fixed 
duties. ^^ Generally, we may say that the universal sex dichotomy 

" Reo Fortune, Sorcerers of Dobti: The Social Anthropology of the Doha Islanders 
of the Western Pacific (London, 1932), p. 15. 

" Bronislaw Malinowski, Crime and Custom in Savage Society ( London and New 
York, 1926). 

^^ W. H. R. Rivers, The History of Melanesian Society (Cambridge, 1914), I, 
60-143; R. H. Codrington, The Melanesians: Studies in Their Anthropology and 
Folk-Lore (Oxford, 1891), pp. 101 f. 


The Family as a Social Unit 

as to occupation is precisely a factor that fosters the family unit, 
for such division of labor, with its frequently correlated part- 
time separation as to companionship, obviously accrues to the 
advantage of the common household. 

One other significant feature must be mentioned as modifying 
family relations. There may be segregation by age or status as well 
as by sex, or both forms of cleavage may be combined. Among the 
Masai of East Africa the bachelors occupy a separate hut, where 
they are joined by the young girls of the village with whom they 
consort apparently ad libitum. This is promiscuity in the popular 
but not in the scientific sense. For with meticulous care the Masai 
abstain from sex relations both with kinswomen and with their 
prospective wives, i.e., girls betrothed to them in infancy. And 
this once more accentuates the persistence of the family concept. 
For notwithstanding the license of the celibates' corral, it is defi- 
nitely expected that every youth and maiden settle down in mar- 
riage after they have had their fling. Premarital freedom is fol- 
lowed by regular family life.^^ 

In other areas, for example, in parts of Australia, only the boys 
are separated from the married couples. Usually this takes place 
after an initiation ceremony, sometimes at the age of seven. Rela- 
tively young boys are thus to some extent liberated from parental 
influence and subjected to the precept and example of somewhat 
older members of their own generation. In Samoa the unmarried 
are segregated from married folk in distinct male and female 
groups. The bachelors cultivate the soil, cook for the masters of 
the several households, and perform necessary communal tasks. 
The female counterpart embraces widows and wives of commoners 
as well as spinsters, and seems to have grown out of the custom 
of having companions of the same age groups and older chaperons 
sleep with a chiefs favorite daughter. ^^ Again, among the Banks 
Islanders the men's club was divided into degrees, membership 
into each being acquired by purchase. Thus, there was a sepa- 
ration not merely of spouses, but of fathers and sons: normally 

" M. Merker, Die Masai: Ethnographische Monographie eines ostafrikanischen 
Semitenvolkes (Berlin, 1910), pp. 44, 84. 

^' Margaret Mead, Social Organization of Manua, Bernice Pauahi Bishop Museum 
of Polynesian Ethnology and Natural History, Memoirs, Bull. 76 (1930), pp. 14, 
92 f. 


Kinship and Social Organization 

a man would eat neither with his wife nor with his children, and 
a mother would be dissociated from her sons as soon as they had 
entered the club house, an act which was rarely deferred until 

I have thus not merely admitted but stressed the diversity of 
family patterns in recent human societies. This differentiation, 
however, virtually 'never militates against the principle that hus- 
band, wife, and child constitute a definite social unit set off from 
other like and unlike units in their community. 

Lest the oddity of some savage arrangements make us lose our 
sense of perspective, it is well to recall historic changes in the con- 
cept of the "family" as held by civilized peoples. Certainly the 
Chinese are not lacking in a family sense, but it is coupled with 
notions foreign to us of wifely duty, of polygyny, and of concu- 
binage. Scriptural patriarchs, too, were polygynous and concupis- 
cent, but no one challenges the prominence of the family in 
Biblical times. Much nonsense is lavished nowadays on the de- 
struction of the family by industrial civilization. Yet the legal 
ties between parent and child, husband and wife, are clearly rec- 
ognized. What has happened is an alteration of the family ideals 
among large portions of our population. For better or worse, the 
change from rural to urban residence, the stress of economic con- 
ditions, an individualistic ideology, the partial abandonment of 
traditional religious doctrines have jointly affected the relation- 
ships involved in the family concept. In the latter half of the eight- 
eenth century Dr. Samuel Johnson, that paragon of Christian 
piety, laid it down as a principle that "wise married women don't 
trouble themselves about infidelity in their husbands." He con- 
sidered a woman who should turn the tables on an errins hus- 
band as "very fit for a brothel." These ideas, I believe, are no 
longer universally held with equal fervor. What I should like to 
point out is that between the upholders of a double standard and 
the modem sex egalitarians the difference is roughly like the dif- 
ference between either and the Murngin or Banks Islanders. Only 
those iconoclasts would fall outside the common practice who 
should consign infants to communal baby farming and who would 
not tolerate any but quite temporary sexual attachments. Such 


The Family as a Social Unit 

societies have indeed been reported with much extravagance of 
vituperation, but with great frugahty of proof. 

A few conclusions of general interest may be summarized: 

1. We know nothing whatsoever about the sex behavior of the 
immediate forerunner of modern hominids except that it very 
probably conformed to the generalized primate norm. Specifically, 
if Mr. Miller's summary is trustworthy, this implies the lack of a 
rutting season. 

2. Though we cannot picture the sexual life of the proto- 
hominid, we may be sure that there was a stage of promiscuity, 
i.e., of socially unchecked sex activity. For, by definition, social 
checks are a characteristic of culture; hence before there was a 
culture there was, in the scientific sense, promiscuity. 

3. All the unequivocally rude tribes of the world — Andamanese, 
Bushmen, Australians, Fuegians, Paiute — have a violent reaction 
against incest with the closest kindred. It is, therefore, extremely 
probable that this sentiment is of great antiquity. 

4. Nevertheless, I no longer believe, as I once did, that incest 
is instinctively objectionable to man. On the one hand, I am as- 
sured on good legal authority that the criminal calendar of Western 
nations shows relatively many instances of paternal lust directed 
against daughters; and if only one tenth of psychoanalytic evi- 
dence is rated valid, the Oedipus complex remains as a factor to 
be reckoned with. As regards siblings, we have at least three his- 
toric cases in which the supposed instinct was deliberately set 
aside — ancient Egypt, Peru, Hawaii. In each of these aristocratic 
societies no mate was considered more appropriate for a ruler 
than his own sister, the only one, evidently, who fully shared his 
illustrious pedigree. 

The aversion to incest is, therefore, best regarded as a primeval 
cultural adaptation which certain individuals potentially or actu- 
ally override in all societies and which certain sophisticated socie- 
ties have expressly disregarded in the interests of an inflated sense 
of aristocratic lineage. 

5. There is no parental instinct. No man can know instinctively 
that he is the begetter of an infant presented by his wife. Demon- 
strably, savage men in many and diverse societies utterly and 


Kinship and Social Organization 

deliberately ignore the question of physiological relationship while 
emphasizing that of sociological kinship. The maternal sentiment 
seems to rest on a firmer basis. Actually, economic pressure or the 
desire to avoid the shame of an illegitimate birth may be stronger. 
Among the Mumgin, "Sometimes a mother kills her newborn babe 
because it has followed too closely to her others and she has not 
enough milk to feed it." Here, as well as in many other savage 
communities, the superstitious objection to twins invariably leads 
to the immediate killing of at least one of them. 

What is of course universal in the interests of group survival 
is a generic interest of adults in children. This sentiment, how- 
ever, as we have just seen, is not manifested by all members of the 
species uniformly, but may be ignored by the superior force of 
utilitarian rationalism or ideological irrationalism. 

6. Every known society distinguishes between mere cohabita- 
tion and that socially approved form of relatively permanent co- 
habitation known as marriage. It may not be superfluous to point 
out that, as there is social fatherhood without the notion of pro- 
creation, so there is frequently social wifehood without phys- 
iological relations. A man may inherit a woman so old that she is 
unfit or undesirable from a sexual point of view; nevertheless, she 
would engage in the feminine occupations with the other women 
of the household and would be entitled to protection and care on 
the part of its master. To cite a concrete case, among the Manyika 
of East Africa a woman becomes the property of her elder sis- 
ter's eldest son. "He does not cohabit with her, but otherwise has 
complete control over her. He may keep her at his kraal, where 
she does the usual woman's work for him. She has no recognized 
husband, but is encouraged to have a lover or even several." The 
children from such unions, it is interesting to add, are in no way 
under the tutelage of their biological father, but are wards of the 
man who inherited their mother; he, and he alone, receives the 
girls' bride-price and provides the boys with the wherewithal 
for acquiring a wife.^^ 

7. Apart from minor modifications or rare and highly localized 
deviations, the family based on marriage is a quite general phe- 

" Charles Bullock, The Mashona, the Indigenous Natives of S. Rhodesia (Lon- 
don, 1928), p. 65. 


The Family as a Social Unit 

nomenon in known samples of Homo sapiens. A man socially 
functioning as a father and husband practically everywhere com- 
bines with a woman functioning as mother and wife to provide 
for their common household and the children begotten by them 
or by legal fiction reckoned as their oflFspring. Since this pattern 
is common precisely to the unequivocally rudest known tribes, it 
is presumably one of great antiquity in Homo sapiens. How far 
back it goes in his history and to what extent it even antedates 
him, no one knows. 



Nomenclature and Social Structure 


light upon certain problems of social organization. My objection 
to using it as exclusively as has sometimes been done is that it 
does not shed enough light. As an introduction to my main argu- 
ment, I can do no better than to quote specific data, as given 



Terms Haisla 

omp (op) father, father's brother, men of 

father's clan and generation 

abe'mp (abu'h) mother, mother's sister, women 
of mother's clan and genera- 

maternal or paternal aunt ane's (ani's) 

maternal or paternal uncle qule' 
(xwatla'p) 1 

father's sister, women of father's 
clan and generation 

mother's brother, male clans- 
man of parental generation 

This paper was found among Professor Lewie's unpublished manuscripts. It was 
written after his retirement in 1950. 

^ Possibly a phonetic equivalent of the preceding term. 


Nomenclature and Social Structure 

The result is striking: Haisla nomenclature is not lineal, but 
"bifurcate merging," like that of so many tribes with a unilateral 
organization. But this is precisely what the Haisla have by virtue 
of their six matrilineal clans. In this respect they sharply depart 
from their southern congeners. According to Boas's latest exposi- 
tion, these have nothing that can properly be called a clan. For 
despite the observed preference for the paternal line in the 
transmission of privileges, automatic patrilineal affiliation of all 
siblings and exogamy within the paternal line are lacking: the 
marriage, reported as orthodox, of a man with his younger 
brother's daughter would be anathema in any typical patrilineal 
clan system.^ 

In other words, that Kwakiutl tribe which has a clan system 
reflects clan affiliation in its kinship nomenclature; those Kwakiutl 
groups which lack a clan system very naturally do not manifest 
any such influence; the one term originally reported as embrac- 
ing "group of supposed common descent" now stands revealed 
as very often applying to outsiders, to mere friends.^ 

These phenomena thus parallel those discovered among the 
Hopi of Arizona, a people differing from near-by fellow-Shosho- 
neans in the possession of both clans and bifurcation with merg- 
ing.* The case differs in that the clan system with its correlated 
features finds its maximum intensity of development within the 
Pueblo area among the aberrant Hopi, while the Haisla scheme 
is clearly but the pale reflection of a Tsimshian pattern. It is 
wholly probable that the clan idea and the associated terminology 
were borrowed in the lump from northern neighbors. 

In this connection I should like to revert to a perennial moot- 
topic. How far does the nomenclature of relationship correspond 
to social structure and usage? I share Kroeber's and Gifford's 
revolt against the intransigent determinism of Rivers; I agree 
that there is a linguistic aspect to the problem that demands 
investigation.^ But I agree in the spirit in which a fair-minded 

^ Franz Boas, Race, Language and Culture (New York, 1940), pp. 360-367. 

' Ibid., p. 368. 

* R. H. Lowie, Hopi Kinship, AMNH, Anthropological Papers, XXX ( 1929), 380. 

^ See A. L. Kroeber, "Athabaskan Kin Term Systems," American Anthropologist, 
XXXIX (1937), 602; E. W. GifFord, "A Problem in Kinship Terminology," ibid., 
XLII (1940), 190 ff. 


Kinship and Social Organization 

biologist of the mechanistic persuasion acknowledges that he 
is unable to reduce life to purely physico-chemical formulae, 
yet insists that life cannot be explained in any other way. I 
contend that while we cannot, and probably never shall, explain 
all terminological features in social terms, it is a case of either 
such explanation or no explanation whatsoever. 

For, deprecating as I do the extreme position assumed by 
Rivers, I cannot but share his perplexity as to what is meant 
by a linguistic determination or interpretation of kinship terms.'"' 
When language subsumes under a common rubric, two phe- 
nomena, whether guinea-pigs and pigs or brothers and brothers- 
in-law, the original labeler evidently must have detected some 
resemblance between the phenomena he grouped together. But 
whatever was the common denominator between the guinea-pig 
and our porker, the association was evidently not inevitable since 
the Spanish equivalent is conejillo de Indias, "little Indian rabbit." 
Now what are the conceivable bases for linking relatives under 
one term? In his justly famous paper first introducing the lin- 
guistic point of view into this discussion,^ Kroeber listed eight 
categories, of which virtually all rest on social groupings. Spe- 
cifically, in accounting for the fusion of the male cousin and 
brother-in-law concepts ( Dakota woman speaking ) , he points out 
that these kinsmen are alike in sex and in being of the sex 
opposite to the speaker's, that they are of her own generation 
and not in her direct line of descent. But sex dichotomy, genera- 
tion, collateral versus lineal descent are one and all sociological 
concepts. In this article, Kroeber differs from Morgan not in 
eschewing sociological interpretation, but in repudiating a spe- 
cific type of sociological interpretation, viz., that which rests on 
forms of marriage. 

The corrective value of Kroeber's and Gifford's several ex- 
positions against easy-going enthusiasm for short-cuts is very 
high, and the same must be said of Aginsky's and Opler's recent 
papers.^ From Aginsky we learn that a classification traditionally 

" W. H. R. Rivers, Kinship and Social Organisation (London, 1914), p. 9. 

' A. L. Kroeber, "Classificatory Systems of Relationship," RAI, Journal, XXXIX 

** B. W. Aginsky, "The Mechanics of Kinship," American Anthropologist, XXXVII 
(1935), 450; M. E. Opler, "Apache Data Concerning the Relation of Kinship 
Terminology to Social Classification," ibid., XXXIX (1937), 201 ff. 


Nomenclature and Social Structure 

associated with special forms of marriage or descent may result 
from the speaker's adoption of either parent's identifications. 
Opler finds challenging disparities in nomenclature between 
Apache tribes sharing the same behavior towards a particular 
relative. In other words, Aginsky demonstrates that the same 
effects may spring from unlike causes; Opler, that identical causes 
seem to yield diverse effects. Nevertheless, we note with interest 
that these authors are very far from abandoning sociological 
causation on principle. Aginsky pertinently asks what affects the 
choice of one or the other parent's identifications and answers in 
terms of economics, residence, inheritance. Opler brilliantly sug- 
gests that terminology is not fully determined for the simple 
reason that a relative may be the focus of a whole series of be- 
havior patterns, some of which may be stressed in one tribe, 
others by a related tribe. For instance, married Apache sisters are 
in many ways socially equivalent, and if this aspect of their 
functions happens to gain emphasis a single word will embrace 
both of them. But sororal polygyny being a possibility rather than 
a frequent phenomenon, there was an equal chance for stressing 
the differences between mother and maternal aunt. 

It is clear, then, that in the light of our present insight a one 
hundred per cent correlation between a specific determinant and 
a terminological correlate cannot be admitted without the most 
rigorous proof. But it is not a case of "all or nothing" any more 
than in other scientific situations. Geographers are not humbugs 
when they tell us that temperature depends on distance from 
the equator even though snow covers the top of Kilimanjaro. 
Quite properly they adduce the "additional factor" of altitude 
to supplement that of latitude. I see no logical objection to follow- 
ing the same procedure in ethnology: it is the reality of the 
factors, not their number that seems significant to me. 

I believe some difficulties disappear if we keep in mind how 
the problems of kinship terminology arise. Morgan found a 
terminology among the Seneca very different from ours and 
subsequently found great similarity between Seneca and Ojibwa 
and even between Seneca and Tamil. What underlies such dis- 
tributional data? The question is a perfectly natural one and 
fairly clamors for an answer. The linguistic "interpretation" would 
come to tliis, that a fortuitous concatenation of psychological 


Kinship and Social Organization 

causes unrelated to social structure and operating on the prin- 
ciple of free association by which some Indians called the horse 
a "deer" and others "a mysterious dog" made the Tamil in India 
and the Seneca of New York classify kinsfolk in a manner so 
similar that an analysis of the one is "nearly a literal transcript" 
of the other.^ On this hypothesis it seems odd that a distant 
Dravidian people should concoct a scheme so much closer to 
the Seneca pattern than that of many fellow-Americans, includ- 
ing even some fellow-Iroquois tribes 1 I must confess that my 
craving for an interpretation remains wholly unsatisfied by such 
a statement. Ridiculous as is Morgan's explanation in terms of 
racial affinity between the aborigines of India and of America, it 
goes at least through the forms of an explanation. 

Now the supposition that similarities in remote areas are due to 
similar causes may be erroneous in the light of Aginsky's and 
Opler's considerations, not to mention Boas's and others' prin- 
ciple of convergence generally, but at least it does not leave us 
completely up in the air. If we single out a particular common 
factor as the cause, we probably overstate the case but are at 
least attempting a proximate analysis. I certainly did overstate 
the case in my earlier treatment of the subject when I blandly 
interpreted the overriding of the generation principle by ex- 
ogamous grouping.^" But I still believe that I was wrong only 
in the sense in which a geographer is wrong who derives tem- 
perature wholly from latitude. Internal analysis of the Omaha 
system shows that when relatives otherwise indistinguishable 
as to proximity, generation, and sex differ in clan affiliation they 
are differently designated. The clan factor does not account for 
everything in view of the differences typified by the Crow, tlie 
Omaha, and the Dakota, but neither can it be ignored: we 
require "an additional factor" corresponding to our climatologist's 

° L. H. Morgan, Systems of Consanguinity and Affinity (Washington, 1871), 
p. 387. 

^° R. H. Lowie, "Exogamy and Classificatory Systems of Relationship," American 
Anthropology, XVII (1915), 238 (reprinted as paper No. 2 in this volume). 

" Idem, "The Omaha and Crow Kinship Terminologies," Verhandlungen des 
XXIV. International Amerikanisten-Kongresses, pp. 102 ff . ( reprinted as No. 7 in 
the present volume). 


Nomenclature and Social Structure 

Mr. Gifford's argument against the sociological interpretation 
of the Omaha system is wholly convincing if (a) that system is 
derived automatically from patrilineal descent; {b) only the 
California data are envisaged. For on the former assumption 
not only the Southern Shoshoneans and Pima (whom he cites), 
but the Ojibwa, the Maricopa, the Baganda, the Jagga, the 
Murngin, the Tungus, ought all to share the Omaha nomenclature. 
And with a restriction to California it would be quite arbitrary 
to pick out the six tribes with patrilineal descent, regarding 
them as the originators of the terminology and the ten clan- 
less tribes as the recipients. 

This, however, does not do justice to the problem as it actually 
arose. What haunted me was the fact that here was a nomen- 
clature utterly distinct from that of fellow- Siouans, who have 
at least two different modes of classification, yet basically like 
that of a solid block of neighboring Algonkians; and these 
again differed from such fellow-Algonkians as the Ojibwa. And 
what obsessed me still more was the amazing fact that nothing 
like this reappears westward of the Southern Siouans until we 
get to central California! That is the cardinal problem. In broader 
perspective the contrary California instances do not weigh against 
the favorable ones in the ratio of 10:6. Rather, all the Calif ornian 
instances represent one distributional block, the Southern Siouans 
and their Central Algonkian neighbors another, such tribes as 
the Assamese Lhota a third. Quite obviously the mere factor of 
patrilineal descent is inadequate; but that rule plus a specific 
form of marriage congruous with it does explain the observed 
distributional data and inclines the balance in favor of diffusion 
from the patrilineal to the clanless Californians in question, a 
result quite properly inadmissible on California evidence only. 

The conclusion to which I am once more drawn is that a 
preliminary to sound inference is a rough knowledge of the 
total world or at least continental range of distribution. This 
naturally and inevitably yields certain problems. When these 
have been formulated, a more refined approach should set in 
with the comparison of groups linguistically and culturally most 
closely related. By concentrating on the points of difference we 
can then venture opinions as to which differences are functionally 


Kinship and Social Organization 

related — in the correct mathematical sense of the term which, 
pace Dr. Blumenthal,^^ does not mean the same thing as causally 

^ Albert Blumenthal, "A New Definition of Culture," American Anthropologist, 
XLII (1940), 576. 


PART II Literature, 
Language, and Aesthetics 

"A Note on Aesthetics" (No. 10) and "A Case of Bilingualism" (No. 
12) are reproduced here in part because they were esteemed by 
Lowie himself. He would probably have also included in this sec- 
tion chapter v, "Literature," of his book The Crow Indians. Since 
the decision was made not to make excerpts of Lowie's books in this 
collection of papers, that chapter has been omitted. In its place, 
although by no means its equivalent, two other papers have been 
selected: "Observations on the Literary Style of the Crow Indians" 
(No. 13) and "Some Cases of Repeated Reproduction" (No. 11). 
The latter is part of his Studies in Plains Indian Folklore that Lowie 
would also have wished reprinted. From both No. 11 and No. 13 the 
Crow texts contained in the original papers have been deleted. If 
any excuse other than expediency must be offered for this excision, 
it is that Lowie's Crow linguistic materials have recently been pub- 
lished by the University of California Press (Crow Texts, 1960, and 
Crow Word Lists: Crow-English and English-Crow Vocabularies, 

A challenging article, "Native Languages as Ethnographic Tools" 
(American Anthropologist, XLII [1940], 81-89), properly belongs 
in this series of papers but has been omitted in view of its availa- 
bility. The final article of Part II, the hitherto unpublished "Evolution 
and Diffusion in the Field of Language" ( No. 14 ) , might have just 
as legitimately been included in Part IV, "Theories and Theorists," 
where Lowie's views on genuine and spurious concepts of evolution 
emerge more fully. 


A Mote on Aesthetics 

While attempting to determine the artistic style of 
Crow parfleches as compared with that of other Plains tribes, 
I hit upon the notion that it might be desirable to apply some 
of the methods in vogue in experimental aesthetics. Circum- 
stances prevented me from carrying these inquiries very far. 
Nevertheless, I feel it may be worth while to record my measure- 
ments in the hope that they may stimulate others to make 
corresponding observations on a larger scale and particularly to 
undertake relevant investigations in the field. 

Gustav Theodor Fechner, the founder of aesthetics as a branch 
of exact psychology, endeavored to determine what forms of a 
particular geometrical category were deemed most pleasing. For 
this purpose he employed three methods, — that of having his 
subjects choose from a series of, say, rectangles the most aesthetic 
samples; that of having them construct the desired forms; and 
that of noting objectively what forms predominated in actual use. 
Since the decoration of parfleches consists for by far the greatest 
number of instances of simple geometrical figures, it seems to 
present an excellent opportunity for applying Fechner's prin- 

American Anthropologist, XXIII (April-June, 1921), 170-174. 



Literature, Language, and Aesthetics 

ciples, though in the study of museum material the first two of his 
methods are of course excluded.^ 

Inquiries of this sort have an ethnographic no less than a psy- 
chological interest. A priori it is indeed possible to assume that 
in respect of the simpler geometrical figures a single aesthetic 
norm is common to all mankind, — say, the principle of the "golden 
cut" examined by Fechner, according to wliich the ideal rectangle 
has sides bearing to each other the radio of 1 d= V 5 to 2, the 
lesser having a length approximately 61.8 per cent of the greater. 
But it is far more reasonable to expect certain differences in the 
aesthetic canons accepted in different regions. And if this an- 
ticipation were verified, we should have an additional set of 
features for differentiating cultures. What is more, by pursuing 
such studies it becomes possible to define existing differences 
with greater nicety: instead of contenting ourselves with the 
remark that one region favors an angular and the other a cur- 
vilinear style of decoration we may succeed in determining ob- 
jectively that one tribe prefers a rectangle of one type, a neigh- 
boring tribe a rectangle of another type. 

But the matter is not quite so simple as this formulation might 
suggest. After one has handled a fairly large number of specimens 
from a single group it becomes clear that the preferences are 
not clear-cut and absolute. For example, we cannot say that the 
Crow use, say, isosceles triangles for the simple reason that 
even the same bag may be painted with right-angled as well as 
isosceles triangles; and the latter again may vary enormously 
in their aesthetic character according to the angle enclosed by the 
equal sides. It appears that the aesthetic value of a simple form 
is affected by its position in the decorative field: what is proper 
in a marginal area may be taboo in the middle, and so forth. 

In order to avoid the pitfalls just hinted at I decided to com- 
pare the parfleches of the Shoshoni with those of the Crow as 
regards a single figure in the same position, to wit, the rectangle 
in the center of the decorative area. The central rectangle has 
been rightly noted as a trait characteristic of the Shoshoni par- 
fleche, though it is by no means found on all Shoshoni speci- 

^ Those interested are referred to G. T. Fechner, Vorschule der Aesthetik ( Leip- 
zig, 1876) and Ch. Lalo, L'Esthetique experimentale (Paris, 1908). 


A Islote on Aesthetics 

mens.^ This feature is to some extent shared by the Crow. That 
it has a single origin historically cannot be doubted considering 
the geographical position of the tribes concerned and the lack 
of this motive on the parfleches of most other tribes. The ques- 
tion, then, is whether the borrowing tribe has transmuted the bor- 
rowed feature in consonance with its own aesthetic predilections 
and wherein such modifications consist. 

So far as I know, the two flaps of a parfleche invariably bear 
the same ornamentation and it is plausible to assume that they 
are meant to be identical. But whatever may be the artist's ideal, 
she frequently departs from it as regards the dimensions of her 
figures. In some instances, indeed, the discrepancies proved 
decidedly startling. I also found that the parallel lines of the same 
rectangle were not always equal in length but sometimes varied 
in appreciable measure. Accordingly, in establishing my ratios 
I measured all the sides and averaged those determining the 
same dimension. Since in the majority of cases there is a frame 
round the central figure, this provided an additional rectangle 
for each flap, so that the number of ratios for any one parfleche 
is usually four. The shrinking of the rawhide and the partial ob- 
literation of some of the lines make exact measurement difficult 
in some of the specimens, but of course the minor inaccuracies 
due to these causes are negligible for present purposes. Only in 
one case were certain lines so completely effaced that measure- 
ment was impossible. 

In the table (p. 140) the fractions designate the specimens as 
registered in the catalogues of the American Museum of Natural 
History. The absolute measurements are given in millimeters, 
those relating to the parallel sides of the same rectangle being 

It would of course be vain to draw any far-reaching conclusions 
from the small number of cases available for comparison. If I 
venture to broach the subject, it is because it provides a valuable 
method for field-workers, which I hope they will not neglect. 
It is not always practicable to purchase large series of museum 
specimens, but few natives would object to having the figures on 

^ A. L. Kroeber, Ethnology of the Gros Ventre, AMNH, Anthropological Papers, 
I (1908), 172. 


Literature, Language, and Aesthetics 








138, 139 




136, 138 


202, 197 

303, 300 



172, 160 

269, 270 



210, 200 

299, 295 


172, 183 

263, 267 


160, 159 

263, 265 



175, 170 

282, 294 



131, 130 

231, 225 


148, 140 

250, 250 


161, 150 

210, 205 



124, 114 

183, 176 



153, 145 

191, 196 


95, 100 

160, 161 


154, 154 

246, 244 



133, 136 

222, 220 



175, 175 

283, 290 


145, 145 

266, 259 



175, 173 

208, 204 



183, 180 

190, 200 







147, 143 

245, 249 



102, 97 

206, 207 



149, 148 

222, 221 


111, 108 

180, 182 


130, 120 

145, 149 



97, 102 

117, 114 



120, 133 

157, 157 


97, 105 

121, 125 


112, 124 

224, 228 



90, 80 

190, 189 



121, 113 

217, 223 


86, 78 

177, 181 



A Note on Aesthetics 

Crow — Continued 































* This specimen was photographed in the field and the proportions calcu- 
lated from the negatives. Owing to the small size of the measurements ob- 
tained, differences between parallel lines are ignored here. 

their rawhide bags (or other objects bearing designs) measured 
by an ethnological visitor. I certainly feel confident that had I 
been alive to this mode of research at the proper time I could 
have secured an imposing array of data on Crow parfleches that 
would have definitely decided the closeness of their kinship with 
those of the Shoshoni. 

I will assume that the samples of ratios supplied by my two 
small series are typical and will collate the data in a table of dis- 
tribution, uniting percentages by fives. 

Ratios Crow 

40-45 1 

45-50 2 

50-55 5 

55-60 3 

60-65 2 

65-70 2 


75-80 1 

80-85 3 

85-90 1 



The fact that the number of Crow cases above 80 is twice 
that of the Shoshoni is readily explained when we remember 
that the Crow piece in question has a framed rectangle while the 



Literature, Language, and Aesthetics 

excessively broad Shoshoni parfleche is frameless. Few as are 
the ratios, all the data consistently point in one direction, — a 
preference of the Shoshoni for relatively wide rectangles in the 
central position. The narrowest Crow rectangle is much narrower 
than the narrowest Shoshoni one; the broadest Shoshoni rectangle 
is broader than the broadest Crow rectangle; the Shoshoni prefer 
quite decidedly the ratio of from 60 to 70, the Crow the ratio of 
from 50 to 60. On the basis of these figures the Shoshoni norm 
would fall somewhat above and the Crow norm somewhat below 
Fechner's ideal rectangle. 

I have already indicated that I attach to these findings a merely 
tentative and suggestive value. Of course comparison should not 
be restricted to rectangles in a particular position but must be 
extended to other forms, say, the diamonds or hour-glass figures 
that are so prominent in the rawhide decoration of Plains Indians. 
A comprehensive inquiry of this sort is bound to yield interest- 
ing results for it will be as important to ascertain that there is 
practical unity of aesthetic reaction to geometrical forms as to 
determine tribal differences. 



Some Cases of Repeated Reproduction 



tobacco society of the Crow Indians. At one stage of the per- 
formance a man with a creditable war record was sent for 
water. After ceremonial preparations he dashed off, filled a ves- 
sel, and returned, thereupon reporting in a very low tone of voice 
to the owner of the initiation lodge. 

Gray-bull, who had repeatedly served in this capacity, three 
times dictated to me the tenor of the water f etcher's (ak'i'cde) 
communication.^ Unfortunately I am unable to give the intervals 
between successive recitations. Nevertheless the variants present 
some points of interest. [Crow texts of the following translations 
are omitted, and their accompanying footnotes.] 


1. On a war party they went, among them I went. 2. The people 
[person?] toward [post-position] they ran, they killed, his gun I 

Studies in Plains Indian Folklore, Univ. Calif. Publ. Amer. Arch, and Etlin. XL 

(1942), 19-28. 

^ R. H. Lowie, The Tobacco Society of the Crow Indians, AMNH, Anthropologi- 
cal Papers, XXI ( 1919), 153 f., 185. I have corrected the orthography in the present 
paper and to some extent die translation. 


Literature, Language, and Aesthetics 

took. 3. Then I came back. 4. The tobacco you [plural] had planted 
when I reached, it was abundant extremely; round about the choke- 
cherries were abundant extremely. 5. Then I came, the camp when 
I reached sick people there were none. 6. Peacefully the tobacco 
you were harvesting. 

1. Chokecherries ripening [i.e., the summer] safely we shall reach. 
2. Sickness there is none. 3. I went on a war party, someone was 
killed, a coup I struck. 4. My heart being good [i.e., happy] I ar- 
rived. 5. The tobacco you planted I saw, the tobacco was plentiful. 
6. The tobacco to see I wished, I came, the tobacco was growing 
excellently. 7, The Crow well were faring, chokecherries in safety 
you were eating. 

1. I went on the warpath, in safety [i.e., without loss] they killed 
someone, his gun I took. 2. Then I came, your gardens when I saw, 
the chokecherries were plentiful. 3. Round about buflFalo were 
plentiful. 4. Then I came, when I came (?) and the camp I reached, 
toward camp I signaled. 5. When I came and the camp I reached, 
the Crow well were faring. 

Evidently the form of the report is not fixed, for tlie same in- 
formant's versions reveal variations. But the substance is identical, 
being composed of three themes: the speaker's war experience; 
his inspection of the tobacco garden and its environment as he 
returns; and his auspicious findings. The emphasis is throughout 
on the rosy side of life: a successful raid; a plentiful crop of 
the weed believed to ensure the tribal welfare; abundance of 
food generally; and the prosperity of the people as foreshadowed 
by the inspection. What, then, is the nature of the discrepancies? 

Apart from purely verbal difi^erences, the initial cue "war 
raid" would present a seasoned brave, such as Gray-bull, with a 
number of possibilities, narrowed only by the need for eliminating 
untoward events. Of the four conventional feats of bravery 
recognized by the Crow, versions (a) and (c) introduce the 
capture of a gun, the other variant the striking of a coup. Since 
Gray-bull himself had in addition successfully led raids and cut 
horses from their pickets, the omission of these feats is not due 
to personal reasons. The failure to mention horses, however, may 
be due to archaizing, since some origin myths made the first 


Some Cases of Repeated Reproduction 

tobacco planting preequestrian.^ Quite naturally a lucky raid 
includes a killing, but characteristically without loss of a Crow. 
This concomitant is understood in the first two versions, made 
explicit by the term i'tsikya'ta in the last. The "signaling to the 
camp" would be out of place in the second version, but would 
be as appropriate in the first as in the third, since acis'-buxu'cuk 
seems to imply signaling that the homecomers are bringing booty. 

Since wild vegetable fare was of subordinate importance, the 
mention of buffalo in a single variant contrasts sharply with 
the appearance of cherries in each version — twice in version (b). 
However, this otherwise curious fact is readily explained: The 
season for the tobacco harvest is "when the cherries are ripe"; 
the phrase ba-'tsua o'-°ce is a cliche for designating the summer- 
time; and like other standardized designations of seasons it ap- 
pears prominently in prayers for long life and happiness. 

The word i'tsikya-'ta, already discussed in a special setting, 
is another cliche. Though requiring different translation accord- 
ing to the context, it invariably denotes a satisfactory situation, 
being simply the adjective i'tsi, good, with an adjectival suflBx 
commonly denoting diminutiveness or affection. Thus, in a 
prayer the suppliant says, "i'tsikya-ta baku' wiawak," In safety 
I want to return; and elsewhere we find, "i'tsikya-'ta bawara'pbic 
bi'awuk," Safely we'll take revenge (compare above). 

The three versions are of some interest as illustrating in 
miniature the type of changes to be expected when the same 
individual reproduces, untrammeled by the necessity of letter- 
perfect repetition, the essence of a traditional pattern. Evidently 
the informant clings to some stereotyped expressions, but he 
has the choice of explicitly stating or merely implying some 
circumstance; he may amplify by adding a specific image (such 
as the abundant buffalo, or the peaceful eating of cherries, or 
the signaling), he may choose between symbols (coup or gun- 
capture ) . 

The foregoing comparison was prompted by Dr. Lindgren's sum- 
mary of Professor Bartlett's experiments on white subjects, each 

' Ibid., pp. 186, 188. 


Literature, Language, and Aesthetics 

of whom reproduced after fifteen minutes, and subsequently at 
various intervals, a North American Indian tale he had read.^ 
It has been recognized for some time that individual storytellers 
within a tribe vary appreciably in their rendering of a tale; and 
this fact evidently affects intertribal comparison. But it seems 
equally important to ascertain how far the same individual de- 
parts from the norm as received by him. This would affect not 
only the evaluation of folk tales, but of "historical" traditions 
as well. 

A good illustration is available. In 1910 I bought a sacred 
shield belonging to Yellow-brow, then in the prime of life, and 
his father. Several years later Yellow-brow gave an account of 
its history, terminating in the description of a battle with the 
Dakota."* In 1931 — without the slightest reference to this shield 
— the same informant launched into a long tradition, which 
similarly culminated in an account of hostilities, but with the 
Cheyenne. The native text, except for one prayer, has remained 
unpublished, but significant passages connected with the battle 
have been presented in English.^ 

Both times the narrator doubtless tried to picture the same 
occurrence. For one thing, the warriors prominent in both ac- 
counts almost entirely coincide in name; for example, Wants- 
to-die, Plays-with-his-face, Double-face, Young-white-buffalo, 
Passes-women. Secondly, both descriptions feature the nervous- 
ness of Double-face before battle, with some identical details, 
such as his desire to cry and to sing both sacred songs and 
those of the Big-dog society. Yet the enemy in the fuller report 
is the Cheyenne, not the Dakota tribe; the entire context differs; 
and along with amazing identities there are notable discrepancies 
not due to mere absence of items in the shorter narrative, which 
as a matter of fact embodies highly characteristic traits lacking 
in its rival. 

I suggest the following explanation. Asked for the history of 

' F. C. Bartlett, M. Ginsberg, E. J. Lindgren, R. H. Thouless, eds., The Study of 
Society, Methods and Problems (London, 1939), pp. 363 ff. 

* R. H. Lowie, The Religion of the Crow Indians, AMNH, Anthropological Pa- 
pers, XXV (1922), 415-418. 

ndem. The Crow Indians (New York, 1935), pp. 230-236, 332-334; idem, 
"Crow Prayers," American Anthropologist, XXXV (1933), 440 ff. 


Some Cases of Repeated Reproduction 

the shield he had sold me, Yellow-brow was primarily con- 
cerned with its validation as an object possessing supernatural 
power. Accordingly, he set out in characteristic Crow fashion, 
deriving it from a revelation in a vision. The visionary, after 
himself profiting from the shield, made replicas for his three 
sons and one nephew, bidding them above all to protect the 
women and children in camp rather than go on raids. Now, in 
his later version the battle follows the destruction of Dangling- 
foot and his small body of Crow by the Cheyenne, whence the 
thirst for revenge on the part of the survivors. Another good in- 
formant, Grandmother's-knife, also ascribed this massacre to 
the Cheyenne.^ But in the earlier of the Yellow-brow accounts 
Dangling-foot is not mentioned; there is merely a generic refer- 
ence to the enemy's having destroyed a detached company of 
Crow. Inasmuch as "enemy" was almost coterminous with "Da- 
kota," the infoi-mant naturally glides from the generic term 
into this specific description, adhering to the identification 
throughout. Even his conclusion, once more summing up the 
case for the value of his shield, is that ever since its acquisition 
the Dakota were repelled by the Crow. 

This seems to explain very simply the shift from Cheyenne to 
Dakota, or vice versa. 

Further, once launched on his glorification of the shield. 
Yellow-brow naturally enhanced its dignity by linking it with a 
stirring tradition of victory that probably had an original con- 
nection with quite different circumstances. However, when simply 
narrating what he regarded as the most striking events in the 
past of the Crow, he enlarged, indeed, on the battle, but the 
shield, however important to him, did not loom in his memory 
in that context, presumably because it did not really belong there. 

This seems a factor to be reckoned with. A mind replete with 
the traditional lore often had alternative sequels for the same 
stage in a story. The lore, for example, may harbor two ways of 
escaping from a pursuing ogre; and a narrator conversant with 
both may choose one or the other according to individual 
preference or momentary caprice. That this is not pure guess- 

* R. H. Lowie, Myths and Traditions of the Crow Indians, AMNH, Anthropologi- 
cal Papers, XXV ( 1918), 185. 


Literature, Language, and Aesthetics 

work is shown by an experience with Yellow-brow when he told 
the Old Woman's Grandson story. He actually told one episode 
in the briefer of two mutually contradictory forms current among 
his people, then corrected himself, retracing his steps so as to 
bring in a prerequisite element for the longer version/ 

Stylistically the sleep-inducing formulae used by Grandson (p. 4), 
and according to some raconteurs also by the snakes he visits, 
are of great interest. The two Hidatsa and my four published 
Crow variants may be supplemented by the following Crow 
versions: an unpublished anonymous Grandson text dictated to 
me probably in 1916; an unpublished Grandson text dictated by 
Plenty-hawk in 1931, which differs appreciably from the printed 
version he had told me before; the three forms of the episode 
introduced, respectively, into the Twin cycle by Plenty-hawk 
and Gray-bull, and into the Old Man Coyote cycle by Grand- 
mother's-knife.^ The total number of comparable formulae is 
thus increased to eleven, which follow. 


(a) 1931 Grandson Text 

1. The rain is dripping fast; we sleep well, don't we? 

2. At the river when we pitch camp, we hear the cicadas (?) 
calling and doze off, don't we? 

3. When it's windy and the tent flaps flap together, we sleep well, 
don't we? 

4. When it is windy and the pine needles rustle, we sleep well, 
don't we? 

(b) 1914 (?) Grandson Myth 

1. Whenever they moved camp by the riverside where there was 
plenty of shade, people would go swimming and in the shade they 
could not help but sleep. 

2. On windy days they came to the tipis and heard the wind 
blowing. Then they would cover up with blankets and could not 
help but sleep. 

3. When a big crowd of people moved toward the mountainside, 

' Idem, The Crow Indians, p. 109. 

* Idem, Myths, pp. 36, 81, 93; cf . with occurrences in the Grandson tale, 
pp. 56, 62, 72. 


Some Cases of Repeated Reproduction 

they would hear a rustling in the pine trees. Then they could not 
help but sleep. 

4. Late in the fall there are long rainy days. We would lie inside 
and put blankets over us and hear the rain strike the tipis, then we 
could not help sleeping. 

(c) 1914 (?) Twin Myth 

1. When a big crowd of people move and reach a river, they are 
always eager to get there. When they arrive, there will be a big 
shade and the river will be high. We'll smell the river and see the 
trees and the leaves floating down and the blackbirds singing over 
the river. After all have camped, everybody will go in for a swim 
and, sitting down afterward in the nice shade, they will fall asleep. 

2. In the fall when the leaves have all turned yellow and are 
falling ofiF, there are sometimes rainy days. They will be out some- 
where far along in the evening and get wet, and when they get home 
they will take a blanket and cover themselves. When they have lain 
thus for a while, they can't help falling asleep. 

3. Late in the fall when the days are windy, they will be out and 
come back home and lie inside. They will hear the wind blowing, 
then they can't help falling asleep. 

4. When they move to the mountains and camp near the pines 
and the wind strikes the trees, they can hear the rustling in the pines 
and can't help falling asleep. 

grandmother's-knife versions 
(a) 1916 Old Man Coyote Tale 

1. When we go along and take a rest under the tall grass and the 
wind moves, we almost fall asleep; and when we do sleep, it's fine, 
isn't it? 

2. It's fine when we are by the riverbank with ripples, isn't it? 

3. Right in the mountains where the streams come out, it's nice 
to lie under the pines and hear the wind blowing through the needles 
and the water running. It is fine, enough to put a man into a dead 
sleep, isn't it? 

4. When the day is cloudy, the thunder makes a low rumble and 
the rain patters against the lodge, then it's fine and nice to sleep, 
isn't it? 

(b) 1916 (?) Grandson Myth 

1. In the fall when it rains, we can hear the rain on the tipi, and 
we shall sleep well. 

2. When we sleep among the pines with the wind blowing and 
we hear the sound of the pines, we sleep well. 


Literature, Language, and Aesthetics 


1. In the spring when cherry and plum blossoms are in bloom, 
when we kill a deer we cook it on the sunny side of a cherry-tree 
thicket. In the fall when it is cool we are out a long time and when 
we come back to our tipi and find it warm we go to sleep right away. 
Do we? 

2. When out hunting in the mountains, when we have killed 
buffalo or deer toward evening and build a fire and cook, while we 
are cooking it grows dark. We are very tired. We take our cooked 
food and eat it. Rain comes and when we lie down to sleep, we 
sleep right away. 


1. In the spring in the daytime when there is a little breeze we 
are wont to sleep well. 

2. In the summer when the raindrops rattle on the tent we are 
wont to sleep well. 


1. In the fall whenever there is a little wind, when we lie in some 
shelter, when dried weeds rub against each other and we listen, we 
generally get drowsy, is it not so? 

2. In the daytime when it drizzles and the rain strikes the lodge 
pattering, we remain lying on the side, and warming our soles, then 
we fall asleep, is it not so? 

3. At night when we are about to lie down, listening to the wind 
rusthng through the bleached trees, we do not know how we get to 
sleep, but we fall asleep. 

4. Having sought a hollow among the thickish pines, we make a 
fresh camp there. The wind blows on us, and we, rather tired, lie 
down and at the same time keep listening to the rustling pines, until 
we fall asleep. 


1. When we move early and camp late in the evening, we usually 
fall asleep and sleep soundly. 

2. On windy days when we do not move and lie in the tipi we 
sleep soundly. 

3. When after being out on rainy days all day we come back in 
the evening and sit by the warm fire, we sleep for a long time. 

4. When we have been out from early in the morning, hunting 
and butchering buffalo till evening, we are tired and go to sleep 
right away. 


Some Cases of Repeated Reproduction 

bear's-arm (hidatsa) version 

1. There is a high butte. You can hear the wind rustling over the 
butte. Then in a moment the rustling ceases as if the butte had fallen 

2. You hear brooks of water making a lapping sound. In a mo- 
ment the lapping stops as if the water had gone to sleep. 

3. You hear the wind blowing, blowing, then all of a sudden it 
dies down just as if it had gone off to sleep. 

4. You hear the leaves of the ti'ces rustling and flapping, and all of 
a sudden you hear no sound any more, just as if those trees were all 
fallen asleep.^ 


1. The first stars you see are still, but toward daylight they always 
shake. When the shaking stops, they are asleep. 

2. When the wind passes bluffs at night, it makes a sound but 
toward daylight it is still, the hills are sleeping. 

3. When the wind passes the Missouri timber, it makes a noise; 
toward daylight it always stops, the timber is asleep. 

4. At night the Missouri River makes a noise; at daylight it stops 
and sleeps. 

These are evidently variations of a single theme with a single 
historical origin. We are confronted with a verbal form of the 
urge to play which Professor Boas has repeatedly stressed in 
other fields of art. It is clear that there are no definite limits to 
variability witliin the same general frame. For that reason a 
clear-cut classification is not easy, apart from the fact that in the 
Hidatsa tales it is the natural phenomena — hills, water, wind, 
trees, stars, Missouri — that fall asleep, not, as in all Crow paral- 
lels, human beings. A further difficulty lies in the fact that only 
three of the Crow versions are available in the original, so that 
the exact equivalence of certain expressions is uncertain. Never- 
theless several points of interest emerge. 

For one thing, eight of the eleven parallels, including both 
Hidatsa forms, are broken up into four sleep-evoking statements. 
It seems highly probable, therefore, that this represents the norm, 
as one might expect from the sacred number of these tribes. 

Secondly, particular informants did not feel bound to adhere 

Martha W. Beckwith, Myths and Ceremonies ( Poughkeepsie, N.Y., 1932), p. 124. 


Literature, Language, and Aesthetics 

to a traditional phrasing. In his Grandson tale, Grandmother's- 
knife duplicates the details of rain pattering against the tipi 
and of the wind heard blowing through the pines, but no other 
features of his Old Man Coyote tale; and he adds the explicit 
mention of autumn. 

Plenty-hawk's three versions agree in introducing the rustling 
pines, the wind, the river, and the rain. But there is wide diversity 
in the ideas evoked by all but the first of these cues. The rain 
may be described merely as "dripping fast"; but in versions (b) 
and (c) it is associated with autumnal precipitation, with the 
comfort of lying indoors covered with blankets being emphasized. 
And while (b) describes the rain heard against the tipi, (c) 
omits the auditory image but explicitly introduces the return of 
drenched travelers to camp in the evening. Moreover, ( b ) simply 
speaks of the fall, which (c) further defines in conventional 
Grow phraseology. 

In all versions the river is linked with pitching camp, but (a) 
adds the call of the cicada; (b) and (c) the picture of shade 
and of swimming; (c) enlarges by explicitly referring to trees, 
to leaves floating downstream, to the singing of blackbirds over 
the water. 

The wind looms large, either implicitly or explicitly. In two 
separate "stories" of (a) it, respectively, makes the tent ears 
flap together and causes the rustling of pine needles which 
latter reappears in (Z?) and (c), (b) not expressly mentioning 
the wind in this context. In ( Z? ) and ( c ) people hear wind blow- 
ing, which is of course implied in (a). 

The differences between the versions by the same informant 
can thus be grouped under two heads: he may expressly state 
or suppress implications ("in the fall" vs. "in the faU when the 
leaves have all turned yellow and are falling off"); and he may 
or may not treat a given situation as a cue for delineating detail 
(river: shade, scenting of river, trees, floating leaves, singing 
blackbirds ) . 

Extending comparison to all eleven, we find that the wind 
figures in every single variant, Hidatsa and Grow, with the 
solitary exception of the Scratches-face story. A river or brook 
appears in both Hidatsa and four Grow versions, these, how- 


Some Cases of Repeated Reproduction 

ever, apportioned between only two narrators. The rustling of 
leaves (or pine needles) seems to be the only other intertribal 
feature, being common to Bear's-arm's Hidatsa story and (ex- 
pressly) five Crow versions (by three informants); interestingly 
enough, this element appears in all three of Plenty-hawk's other- 
wise appreciably varying narratives. 

Uniformly present in all Crow versions, but absent from the 
Hidatsa, is the rain element. The most frequent image (five 
times) associated with it is that of the drops striking against 
the tipi. Next in frequency are an autumnal scene (five versions, 
four informants) and the comfort of being warm indoors (five 
versions, four informants). That comfort may be effectively ex- 
pressed by contrasting it with the wind or rain or cold outside 
(Plenty-hawk [b] and [c]; Scratches-face; Gray-bull); and the 
sense of coziness may be suggested either by the warming of 
one's soles (Yellow-brow), or by more general references to a 
fire and warmth indoors (Scratches-face), or to rolling up in a 
blanket (Plenty-hawk [b] and [c]). 

The fascination of these parallels lies in the fact that we are 
here privileged to catch the folk imagination at work. These 
Indians are not automatically transmitting a fixed form, even 
though of course a few phrases have become stereotyped. In- 
stead, there is the tradition of a generic pattern (sleep-inducing 
conditions) with a few subpatterns (fatigue, shade, specific 
sounds), which the individual is free to develop as he lists, 
so that the unique features of the several variants are, in their 
totality, impressively numerous. What the more elaborate of the 
versions strive for, evidently, is vividness of description to be 
achieved by the accumulation of sense impressions, so that we 
get veritable imagist poems in miniature. The images, however, 
are not by any means preponderantly visual, but to a very strik- 
ing degree include auditory and to a somewhat lesser extent 
kinesthetic ideas. 



A Case of Bilingualism 


been described at length. Accordingly, the following notes, 
though inadequate, may prove of some interest. They are un- 
avoidably autobiographical, for which I apologize. 

I was born in Vienna in 1883, My father was a Hungarian 
from the vicinity of Stuhlweissenburg, southwest of Budapest. 
In that section of the country German had remained dominant, 
so that he learnt Magyar as a foreign tongue. My mother was 
Viennese, and, accordingly. High German was the language of 
our household. My father's was a generaHzed South German form, 
my mother's richly flavored with the racy vernacular locutions 
which even educated Austrians affect. Typical are such words 
as Bissgurn ("termagant"), dalket ("awkward, gauche"), hopa- 
tatschet ("supercilious"). She was capable of expressive original 
creations, such as verhallipanzt ("entangled, confused"), which 
appears in no Idiotikon Vindobonense I have been able to 
consult. Again, like many educated Austrians, she was somewhat 
easy-going on certain points of grammar, substituting the dative 
for the genitive with wdhrend and wegen. On the other hand, 
her father, a physician, austerely criticised such derelictions when 

Word, I (December, 1945), 249-259. 


A Case of Bilingualism 

I indulged in them. It was lie, too, who urged his daughter to 
keep up her children's German in America since we were likely 
enough to learn English there. 

When we left Vienna to join my father in New York, where 
he had preceded us by three years, I was ten and had just 
passed the entrance-examination for a Gymnasium, my sister 
being two and a half years younger. We immediately entered 
public schools and rapidly acquired fluency in English. My 
mother, obeying her father's injunction, maintained German as 
the sole medium of communication between parents and children, 
though my sister and I soon came to speak to each other more 
frequently in English. The family intimates were all Austrians 
and Germans, and though our morning newspaper was English, 
in the evening and on Sundays we regularly bought the Staatszei- 
tung. The Sunday edition of that paper had a puzzle-column, 
over which we pored for hours, winning several prizes in the 
form of German books. We occasionally went to the two German 
theatres and in later years visited German societies. We read 
the classics and the serial modern novels that appeared in our 
Sunday Staatszeitung. 

Nevertheless, our German could not possibly develop as it 
would have in Austria. The range of topics discussed with our 
parents and their friends did not coincide with that thrust upon 
us in the classroom and in association with age-mates. It was 
not as a matter of course, but through later deliberate effort, 
that I learnt gleichschenkliges Dreieck, Herrentiere, and Be- 
schleunigung as the equivalents, respectively, of "isosceles tri- 
angle," "primates," and "acceleration." Similarly, dealings with 
storekeepers were largely in English. Important, too, was the fact 
that there were, of course, no compulsory school compositions to 
be scrutinized by the Argus-eyes of a German pedagogue. 

However, the impulse to use our native tongue creatively re- 
mained strong. My sister wrote original poems in it and translated 
at least one lyric by Tennyson; I translated one of Washington 
Irving's sketches and a passage from Dr. Johnson's The Vanity 
of Human Wishes. Indeed, I made my debut in print at fourteen 
and a half in the New Yorker Revue with an article on Edgar 
Allan Poe, and my first earnings as a writer came from an essay 


Literature, Language, and Aesthetics 

on Ernst Haeckel in the Staatszeitung in 1901. The latter was 
reprinted without my permission in the anniversary volume 
dedicated to Haeckel in 1914.^ I mention the fact because its 
appearance in Germany indicates that it was grammatically ac- 
ceptable, though adolescent in style and sentiment. Incidentally, 
in token of our linguistic conservatism, I note that down to the 
present we prefer writing German in "Gothic" script, as our 
mother had always done. 

In point of vocabulary my German, as explained, lagged be- 
hind my English in various respects, yet it remained ahead of 
it in the domain of domestic utensils and the like. "Skillet," 
"rolling-pin," and "saucepan" still click less immediately in my 
consciousness than Bratpfanne, Nudelwalker, and Reindl (Aus- 
trian ) . 

Facility in German composition, of course, implies much more 
than lexical knowledge; it means, among other things, a control 
of stereotyped phrases, such as Beziehungen pflegen, Possen 
reissen, Nachruf auf. . . This is one respect in which the emigrant 
is handicapped; he knows them, but they are not always at his 
beck and call; hence, at a pinch, he falls back on a correct 
enough, but vaguer, colorless expression which a stay-at-home 
of equal cultivation would spurn. 

Grammar presented difficulties of its own. The Austrian ver- 
nacular, for example, tabus the imperfect, which it supplants with 
the perfect. Hence the correct forms of the preterite were matters 
to be learnt from reading, not through conversational osmosis. 
Then there are some regional differences as to gender: no 
Viennese spontaneously says der Schinken, but die Schinke. 
Again, perfectly familiar nouns are not likely to be declined often 
in the ordinary household routine, hence doubts arise concern- 
ing weak and strong forms, and den Hirschen may usurp the 
part of den Hirsch. Thus, eternal vigilance is the cost of maintain- 
ing tolerably good German in a foreign country. We achieved 
the satisfaction of having our German pronounced much better 
than that of other children among our acquaintances. 

Two vitiating influences had to be specially contended against. 

^R. H. Lowie, Was wir Ernst Haeckel verdanken, ed. H. Schmidt (Leipzig, 
1914), 11,404-407. 


A Case of Bilingualism 

On the one hand, an Auslandsdeutscher rarely resists the temp- 
tation to interlard his speech with alien words. The extreme ex- 
ample is Pennsylvania Dutch. A. R. Home's Pennsylvania German 
Manual for Pronouncing, Speaking and Writing English (Allen- 
town, Pa., 1896) offers a scene from Hamlet in the vernacular. 
The ghost reveals himself in these words: "Ich bin deim dawdy 
si shpook," and the entire passage is of the same order. But 
the New York German of the less educated immigrants was 
not much better in the eighteen-nineties. At the Germania 
Theater I saw a play, "Der Corner Grocer von der Avenue A," 
whose dialogue was of a piece with the title. But even cultivated 
people lapse into the use of such convenient words as "ice-box," 
"car," "elevator." Against this tendency I developed a violent 
distaste, so that in my early twenties I joined the newly founded 
New York branch of the Allgemeiner Deutscher Sprachverein, 
which attempted to purge the language of loan-words. 

This meant that in extending my English vocabulary I re- 
mained restive until I had learnt the proper German equivalent. 
This was a formidable undertaking, for, especially when travel- 
ing in the West and in Canada as an ethnographer, I encountered 
many things that did not exist in Vienna as I had known it or 
among New York Germans, e.g., stern-wheelers and narrow-gauge 
railroads. It gave me a peculiar thrill to learn that they were 
rendered Heckraddampfer and Schmalspurbahn, and to the 
present day a never-failing source of pleasure has been to dis- 
cover such terms as Durchschlag ("carbon copy"), Kotfiigel 
("fender"), and umschalten ("to shift gears"). 

A still more serious, because subtler, peril than the intrusion 
of English words lies in the spontaneous, unsuspected transfer 
of English idioms and the misuse of German words because of 
English models. I once used nur instead of erst for "only," and 
on another occasion spoke of having vermisst (instead of ver- 
passt) a train. Similarly, an Austrian lady wrote about her 
Bente when she meant Mietzins, and nothing seems more natural 
than to aufrufen someone on the telephone when usage demands 
anrufen. Lapses of this order always left me with a sense of 
shame, even when I myself discovered and corrected them. 

Still another subtle influence is phonological: the simultaneous 


Literature, Language, and Aesthetics 

use of similar, but not identical, phonemes in two diverse forms 
of speech tends to affect pronunciation of the sounds in the 
medium less frequently employed. On this interesting subject I 
am able to offer only a single observation. My r came to merge 
with the American r after and before consonants. It required 
deliberate remedial effort to restore the linguo-apical rolling, 
so that about fifteen years ago the late Professor Prokosch at 
Yale specially commended my pronunciation of this phoneme. 

My German was further modified by contact with many 
speakers who were not Viennese. Of course, even during my 
childhood I had occasionally heard North Germans; in fact, one 
aunt was married to a Mecklenburger. However, such experi- 
ences with alien dialects were sporadic. In New York, on the 
other hand, they were repeated and constant. Our landlord and 
his wife were Low Germans from Westphalia, the shoemaker 
was a South German, the tailor a Bohemian. More important, 
our intimates included two former university students from 
the Reich, a Berlinese pharmacist, and a Badener of noble 
descent. Occasional visits to the principal literary society also 
meant hearing a variety of accents, if not dialects. These in- 
fluences modified my pronunciation, my vocabulary, even my 
grammar. I came to lengthen the vowel in Mond and to waver 
between making a terminal g (e.g., in verniinftig) a stop, as 
in Austrian, or a fricative, as in stage-German. I picked up 
unfamiliar words, like hanebiichen "coarse." I dropped even in 
oral communication the Austrian-Bavarian contempt for the geni- 
tive in favor of the dative (dem Voter sein Hut for der Hut des 
Voters). I made increasing use of the colloquially lacking pret- 
erite.- My German, in short, became synthetic. 

I had no opportunity to return to Europe until I was forty-one, 
when I spent about two and a half months in Germany and 
Austria. Six years later I visited these countries again for a 
somewhat shorter period. The principal cities in which I stayed 
were Vienna, Berlin, Hamburg, Munich, Nuremberg, and Weimar. 

The essential facts, then, are that I have preserved a deep 
devotion to my mother-tongue throughout my life, but after my 
tenth year, except for a few months, have lived far from a one 

^ Cf. J. M. Lutz, Baijerisch (Munich, 1932), pp. 28-78. 


A Case of Bilingualism 

hundred per cent German-speaking atmosphere. Especially since 
my removal to California in 1921 and before the advent of 
refugees from the Nazi regime, oral practice in German was 
very rare. It now remains to assess the resultant of these factors. 

This question is not a simple one. There are many grada- 
tions of linguistic proficiency, and this is likely to differ markedly 
in writing and in speech. If the touchstone is indistinguishability 
from native usage, we must remember that not all native judges 
are equally perceptive, and that their conclusions may differ 
according to whether they listen to the subject when he speaks 
his vernacular or when he is deliberately transcending it. 

Summarizing my own feelings as well as many comments 
passed on my oral German in the course of years, I should 
formulate the matter as follows. Naive, uneducated Germans 
generally accept me as a German of unspecified origin, for they 
cannot readily conceive equal fluency on the part of a foreigner. 
Sophisticated observers detect the Austrian flavor and the syn- 
thetic quality of my High German, and tend to identify me as 
an Auslandsdeutscher, as did an educated man with whom I 
fell into conversation in Hamburg in 1924. In answer to a query, 
he declared definitely that he had recognized me as one to the 
manner born, though exposed to alien influences. This accords 
with my own feelings: I have in the course of my travels met 
many Swedes, Netherlanders, Czechs, and others who spoke 
fluent and generally correct German, but hardly ever without 
a sense of my superiority in the use of the language. Naturally 
even my most stilted High German never suggests a Northern 
accent to the discriminating; on the other hand, my Viennese 
has been pronounced authentic by those who ought to know. 

Writing, of course, is a different matter. It seems best to 
discuss this topic later when comparing my proficiency in German 
and English, respectively. 

I was not a complete novice in English when we arrived in 
New York, since I had for several months taken lessons from 
an Austrian lady. To be sure, I had not got very far. Her pro- 
nunciation was hardly exemplaiy, for I pronounced ten with a 
long vowel and had no end of trouble with the quality of the i 
in kill. Doubtless, too, she followed the Gennan notion that u 


Literature, Language, and Aesthetics 

in but corresponds to German o.^ However this be, rapid as- 
similation o£ English at once became a primary goal in America. 
I made it a rule to learn at least ten new words a day. Toward 
the end of my grammar-school days I conceived an inexplicable 
admiration for Dr. Samuel Johnson, whose Rasselas I read and 
on whose style I patterned my own with results easily imagined. 
My teacher in the highest grade repeatedly had to censure my 
passion for sesquipedalian words. Folklore develops easily among 
children; I was soon dubbed "the fellow that swallowed the 
dictionary," and on one occasion I overheard myself described as 
having gone through the tome as far as H! In imitation of the 
great lexicographer I even started a rival volume that was to 
illustrate the use of difficult words by quotations from the 
English classics. 

By the time I graduated from public school my spoken English 
was superficially not perceptibly different from that of any 
thirteen-year-old New York boy. Closer inquiry would have 
established then, as now, the deficiencies already in part alluded 
to: only a New England wife made me realize the true essence 
of a "saucepan"; I never encountered the phrase "milling around" 
until I was on the staff of the American Museum of Natural 
History; and within the past year I spoke of somebody's being 
"the split image" (instead of "the spit and image") of someone 
else. When colleagues credit me with an exceptionally wide 
vocabulary, I therefore feel bound to qualify the comment. I 
know many long and unusual words, but I am ignorant of common 
locutions and not sufficiently conversant with everyday words. 
In lectures and academic discussions I am fluent enough, but in 
recounting a simple occurrence of daily life I am likely to grope 
and fumble for the mot juste — say, "running-board" or "dustpan." 
I constantly marvel at the racy oral English of monoglot New 
England narrators of moderate education and feel that their 
achievement is utterly beyond my reach. Incidentally, interloc- 
utors have often chided me for a certain pomposity in speech. 
In my opinion, this is largely due to my not having the ap- 

^ "Deutsche miissen vor allem vor der schauderhaften, gangbaren Schulaussprache 
mit dem b-Laut gewarnt warden," Otto Jespersen, Lehrbuch der Phonetik ( Leipzig 
and Berlin, 1904), p. 156. 


A Case of Bilingualism 

propriate colloquialism at the tip of my tongue, so that I am 
driven to seek refuge in a colorless blanket or bookish term. 

In apparent conflict with my admiration for the homely au- 
thenticity of English speech as spoken by some Englishmen 
and Americans stands my linguistic authoritarianism. Intellec- 
tually I recognize, of course, that "standard" forms are factitious; 
emotionally I resent deviations. I automatically rank British 
above American usage and at times wonder at neologisms such 
as some scholars freely indulge in — say, Kroeber's "formulable," 
"authenticable." I am shocked by Sapir's defence of accusative 
"who" and outraged by his repeated use of "nuanced" as though 
there were a verb "to nuance." ^ Incidentally, a one-time disciple 
of his calmly speaks of "sciencing." 

Probably because of my bilingualism I do not relish even 
wholly legitimate latitudinarianism, such as Jespersen prizes as 
a signal virtue of English. I wish "people" and "committee" were 
always used with either singular or plural verbs; that a horse 
were not alternately "it" and "he"; that one could not refer to 
mankind as "they" {Oxford Dictionary) or "it" (common usage) 
or "he" (Elliot Smith, Rivers). 

My pronunciation is not distinctive so far as I can gather, but 
long residence in the West sporadically affects my r's, which I 
sometimes suppress and sometimes roll. New words, especially 
proper names, I occasionally accent erroneously, e.g., "Ha'bak- 
kuk" for "Habbak'kuk." My wife notes that in comparatively in- 
frequent phrases I tend to stress the noun instead of its sub- 
stantive modifier, e.g., "home week'" instead of "home' week." 

The relative attitude of a bilingual toward his two media is 
not easily determined. Albert Schweitzer, an Alsatian, has offered 
some valuable hints, which partly but not wholly coincide with 
my observations.^ I agree that it is self-deception if anyone 
describes himself as having two mother-tongues, i.e., as being 
able to shift with absolute perfection from one medium to the 
other in all situations. He suggests a test which no claimant he 
has known was able to pass: computing with equal faciHty in 

* A. L. Kroeber, Configurations of Culture Growth (Berkeley, 1944), pp. 79, 106, 
226; Edward Sapir, Language (New York, 1921), pp. 103, 166. 

° Albert Schweitzer, Aus meinem Leben und Denken (Leipzig, 1932), pp. 51 ff. 


Literature, Language, and Aesthetics 

both languages and designating with equal spontaneity kitchen- 
utensils or the tools in a carpenter's kit. 

In my opinion, however, the phenomenon is more complex 
than that. Thus, I find no consistent dominance of either English 
or German. As the preceding pages show, I have actually em- 
ployed English far more than German throughout the last half 
century, but I do my sums more spontaneously in German — in 
fact, until I had to teach school it never occurred to me to do 
arithmetic otherwise. If I wrote poetry, I think it would be in 

In reading professional prose I can detect no difference. 
As to fiction, a test is difficult because one cannot readily find 
works of equivalent character, hence of similar range of vocab- 
ulary. A modern novel, like Heinrich Mann's Die grosse Sache 
(1930), hardly contained unfamiliar words apart from such slang 
terms as tiirmen ("to scram"), neppen ("to overcharge"), doof 
("stupid"), pennen ("to sleep"), Nutte ("whore"). But, then, 
what American reader knows (not merely guesses from the con- 
text) the meaning of "harlican," "clacker," "tassets," "chitterling" 
in Thomas Hardy's Jtide the Obscure? 

I rather think that I have a deeper understanding of German 
poetry, but here again it is extremely hard to make a fair com- 
parison. The English poet I have read most in recent years is 
Browning, whereas in German I have reverted to Goethe and 
sporadically to lyricists of the Theodor Storm period. 

In writing or lecturing, my facility is certainly greater in 
English, for I have literally written thousands of pages in it and 
lecture six times a week besides conducting a seminar. For a 
lecture of any consequence in German, e.g., when I addressed 
the Anthropologische Gesellschaft in Vienna in 1924, I alwavs 
carefully prepare, so as to preclude the groping for proper words 
and phrases. However, I found it easy to discuss at least one of 
the papers read at an international congress in Hamburg in 
1930. In ordinary conversation I am likely to be equally halting 
in both tongues when it is a question of describing a connected 
series of concrete happenings. I grope for the exact word required 
and grow embarrassed; or I seize upon an unusual, hence stilted, 
mode of expression. Amnesia concerning words afflicts me in 


A Case of Bilingualism 

either tongue. A priori I suppose that, on the whole, it is more 
frequent in German, but I recall one instance when I vainly 
tried to conjure up English "puddle" and got it only after recall- 
ing three German equivalents — Tiimpel, Pfiltze, and Lache. 

To sum up, I am impressed with the difficulty of mastering a 
single language, let alone two languages, in the fullest sense. 
Perhaps the situation is different in such countries as Switzer- 
land, but Schweitzer's Alsatian parallel suggests the contrary. 
Conrad Ferdinand Meyer wrote his masterpieces in German, 
Turgenev his in Russian, though both had admirable control of 
French. Rolvaag's novels were first printed in English, but were 
all composed in Norwegian, though almost all of the author's 
adult life had been spent in the United States. 

There is another aspect of the problem to be considered. A 
bilingual is the linguistic sample of the sociologist's "marginal 
man." He suffers in his use of either tongue when judged by 
the highest standards, but by compensation he has insights 
not granted in quite so vivid a manner to others. He cannot help 
constantly comparing modes of expression; and what others 
recognize as an abstract principle is to him an ever-recurring 
vital experience — the incommensurability of different languages. 
This is, of course, obvious when the speakers have evolved differ- 
ent cultural traits, say, the German Schoffe or the English sheriff. 
But far more significant is the fact that English does not dis- 
tinguish between the bleat (Bloken) of a sheep and the Meckern 
of a goat; that gonnen has to be paraphrased "not to begrudge"; 
that Schiitzling can be rendered only by the loan-word "protege." 
On the other hand, why does German lack a designation for 
"understatement" or even for so common an action as a human 

The popular impression that a man alters his personality when 
speaking another tongue is far from ill-grounded. When I speak 
German to Germans, I automatically shift my orientation as 
a social being, I spontaneously adapt myself to the atmosphere 
characteristic of their status, outlook, prejudices. The very use 
of the customary formulae of politeness injects a distinct flavor 
into the conversation, coloring attitudes and behavior. Some of 
these modes of expression, to be sure, are merely meaningless 


Literature, Language, and Aesthetics 

formulae, but by no means all. The retention of titles, in Euro- 
pean fashion for example, colors mutual relations, as does the 
free and easy American way of dropping them altogether. In 
this respect I find myself still strongly under the spell of early 
influences: whereas my colleagues Kroeber and Olson at once 
call our students "Jones" or "Cooper," I keep up the "Mr." in- 
definitely. Similarly in contact with fellow-members of the faculty 
I am probably never the first to drop the title. 

As a marginal man, in short, the bilingual is something of a 
problem child. It is no surprise to me that "bilingualism has an 
adverse effect on achievement in intelligence tests." ® Language 
is so intimately interwoven with the whole of social behavior 
that a bilingual, for better or worse, is bound to differ from 
the monoglot. His attitude in a novel environment has been 
forcibly depicted by Rolvaag and, I think, still more poignantly 
in the following poem by my sister, Miss Risa Lowie, who per- 
mits me to publish it: 

The Foreigner Speaks 

You wonder at my diffidence, my wistful smile — 
You that never had Whither and Whence to reconcile. 

You wonder — whose vernacular rings like true coin, 
Whose Present and whose far Past are fields that adjoin — 

Why I must ponder simplest things: It must be so — 
Twice I must cross the sea on wings for each Yes or No — 

Appraising at my far Exchange your casual word, 
To see it soar to values strange, or drop absurd — 

Returning, amber in my hands, and driftwood blue — 
Both gathered on my native strands: but which for you? 

" Klineberg, "Mental Testing of Racial and National Groups," in H. S. Jennings 
et al.. Scientific Aspects of the Race Problem (London, New York, Toronto, 1941), 
pp. 260 ff. 



Observations on the Literary Style 
of the Crow Indians 


In one of his earlier papers Professor Thurnwald 

Vor allem mochte ich . . . auf die grosse Verschiedenheit der 
Begabung und des Charakters unter den Einzelnen hinweisen. Es ist 
eine sehr verbreitete Ansicht, dass die Charakter-verschiedenheiten 
unter den Individuen bei den sogenannten Naturvolkem sehr gering 
sind. Das mag dem fliichtigen Beobachter so erscheinen. Wer aber 
nicht nur langer mit den Leuten zusammengelebt, sondern sie auch 
aufmerksam beobachtet hat, dem werden sich die Unterschiede 
zwischen den Personhchkeiten wie die unter den Physiognomien 

This variability appears very clearly w^ith respect to aesthetic 
capacity. As Thurnwald incidentally states in a later publication, 
the Buinese hire "professional" poets to compose songs for them.^ 

Beitrdge zur Gesellungs- und Vdlkerwissenschaft Festschrift Professor Dr. Rich- 
ard Thurnwald zu zeinem achtzigsten Gehurtstag gewidmet (Berlin, 1950), pp. 


^ Richard Thurnwald, "Im Bismarckarchipel und auf den Salomoninseln 1906- 
1909," Zeitschrift fiir Ethnologic, XLII (June, 1910), 146. 

^ Idem, Profane Literature of Buin, Solomon Islands, Yale University Publications 
in Anthropology, No. 8 (1936), p. 6. 


Literature, Language, and Aesthetics 

In the present essay I wish to offer two versions of the same 
historical event, possibly dating back seventy years, and to il- 
lustrate by them the individual variability of Crow narrative style 
as well as features that characterize the tribal pattern. 

The events described revolve about a highly typical institution, 
shared by some neighbouring tribes. Although every Crow was 
expected to be brave, certain men who for some reason had lost 
all interest in life would deliberately court death. Assuming a 
distinctive costume and other regalia, such a one would appear 
in camp, singing a song of his own, saying the exact opposite of 
what he meant, and expecting others to speak to him in the 
same fashion. Thus, to make him dance it was necessary to bid 
him refrain from dancing. By such behavior he gave public notice 
that he was a "Crazy Dog" ( micgye-wara-'axe ) intent on being 
killed by the enemy during that season. A man of this calibre 
was highly honoured, but he lost face completely if he tried to 
annul his promise or evinced cowardice. On the other hand, 
no ignominy attached to a Crazy Dog who honestly tried to 
carry out his resolve, but somehow failed of achieving it. Such 
a one was privileged to renew his pledge the following "season," 
i.e., during the annual period of tribal reunion following the 
dispersal into minor groups during the winter. For instance. 
One Horn's father-in-law was killed during his second assump- 
tion of the part. 

The motives for becoming a Crazy Dog varied. In the in- 
stance just cited the pledger was dissatisfied with the United 
States Government Agency's distribution of rations. Spotted Rab- 
bit was disconsolate because of the death of his stepfather, Good- 
Crazy-Dog ( recte Strikes-the-Enemy-in-His-Brother's-Company ) 
when the Sioux killed some of his relatives. Young Cottontail- 
Rabbit, the hero of the following narratives, could not get over 
the handicap of a shattered kneecap, which prevented him from 
joining in the forays of his age-mates. Naturally there were 
never many men simultaneously bent on terminating their career. 
The largest — quite probably apocryphal — number ever men- 
tioned was ten; one informant, One-Horn, remembered as many 
as five at one time. The usual number seems to have been two, 
and some years no volunteer was forthcoming. According to Hol- 


Literary Style of the Crow Indians 

man, Knocked-over-by-the-Deer ( uux-aripu'cc ) and Many-Butter- 
flies ( Biri'kyac-aho-'c ) were contemporary with Young-Cotton- 
tail, but the former got frightened and quit. 

My first version was dictated to me in 1907 at Crow Agency 
by No-Shinbone, to whom I also owe a graphic description of 
one of his war parties, as well as other information. The second 
variant was obtained in 1931 from Yellow-Brow, admittedly one 
of the best storytellers. 


[Crow text and accompanying footnotes omitted] 
1. At the Old Agency there goods were being distributed, a 
Crazy Dog (I then) came to know for the first time. 2. People 
were seated before the distribution when a young man came 
riding, holding his blanket close to his stomach and making a 
rattle of his quirt. 3. Then he entered the camp circle and now 
I saw the Crazy Dog. 4. Then he sang coming into the circle. 
"Who is it?" people asked. 5. The young man's knee was swollen, 
he had been shot, he envied others, he wished to die, that's why 
he acted thus. 6. Then for a while we never saw him. 7. Then 
one evening he put on his regalia and came splendidly. The whole 
camp, we people, were most eager to see him. 8. Then for his 
rattle, of one of these baking-powder cans he made it, inside he 
put beads, it rattled mightily. 9. The bridle of his horse had fine 
chains (?), he put it on, his horse could not be seen because of 
its trappings. 10. He came, holding his gun by his waist. (For 
a wrist-band) he put the light-colored end of a tail on his rattle. 
11. He himself made his queue, he made little forehead braids 
for himself, he made himself ear-rings. 12. He put shell on his 
ear-rings; a necklace he put round his neck, it was exceedingly 
handsome. 13. His bay horse was bald-faced, it pawed the ground 
mightily. 14. We saw him, the whole camp liked him. 15. Then 
he went through the camp, he came singing, he came shaking his 
rattle. 16. We did not know that he spoke cross-wise. 3 17. He 
came on. A man said (to him), "Don't dance!" He dismounted 
at the entrance of a lodge. Then the young man's drummer sang; 
holding a drum like this one, he sang. 18. He danced. "I'll test 
myself, I want to die. Whether it will come true or not, I shall 
know." He shot down at his foot. "It will be well," he said, it is 
said. 19. The women liked him exceedingly. Every evening he 
danced. The Crow moved, he sang. When the people had all 
^ That is, by contraries. 


Literature, Language, and Aesthetics 

camped, he went singing through the camp. The old women 
cheered him mightily. At night he always sang. 20. When the 
people wished to hunt, they regarded him as a dog.^ He put on 
his sashes and mounted his horse. All the men went hunting, they 
wanted to kill buffalo, with arrows they wanted to hunt. 21. The 
Crazy Dog went hallooing. When these dogs see cattle,^ they 
are wont to run towards them, that is the way he acted. 22. They 
killed many buffalo, they butchered, they packed their horses. 
The people pitched their tents, he went singing through the camp. 
Then the next morning they moved, in a dry gulch with sloping 
sides we camped. 23. A young man had lost his horse, he turned 
back, he went to the old campsite, on the site he saw the enemy. 
He fled, he fell off his mount, the horse ran away. Afoot he reached 
the camp. "The enemy is in the old site," he said, it is said. 24. 
Then they (the Crow) charged them. Several young Crow fought, 
they (the opponents) wanted to kill one another. 25. They drove 
them (the enemy) back, within a creek-bed they (the enemy) 
put up defensive works. The Crazy Dog got there, he wished to 
die. He got to the edge of the defensive works, he shot into them, 
then they killed him. 26. Then it rained violently. The Crazy 
Dog was lying in the water. At night he lay there until daylight. 
27. The next morning we came there, he was lying in the water. 
They made wrappings, they brought a horse, they took hold of 
him, they packed him on (the horse), they led him (the horse). 
They cried as they were going along, they brought him to the 
camp. The whole camp cried very much. 28. They planted sticks, 
there they laid him. They planted a four-pole scaffold and a 
lodge-pole, to which they tied his sash. Moving without him, we 
left. That is how he was killed. His drum — like this one — they 
hung up where they buried. This is the end. 


[Crow text and accompanying footnotes omitted] 


1. He had a father, he had a mother. Then his mother died; 
now he was motherless. It was with his father that he lived. 2. Then 
his father took a wife; Young Cottontail Rabbit there lived then. 
3. This father of his had two children, female ones. 4. Then again 
he married another woman; the woman, too, had children, two 
boys. 5. He (Cottontail) went to war; about that I don't know 
exactly, but he was about eighteen years old. 6. Then Twitching- 
Eyes was the captain. Way yonder on the other side of the moun- 

* That is, they would speak to him as one would in shooing off a dog. 
^ The Crow word is based on the stem for 'TDuffalo." 


Literary Style of the Crow Indians 

tain they were badgered (by the enemy). His (Cottontail's) knee 
was pierced with an arrow, the arrowhead did not come out of 
the kneepan, it was visible, 7. Then they could do nothing, he 
tried to pull it out, he failed. 8. People came, his leg was swollen, 
there was nothing like it.^ 9. Now they were unable to do any- 
thing. "Now, you cannot help me at all, go without me." 10. They 
made a shelter, they got a great deal of firewood for him, every- 
where inside (the shelter) they himg up water-paunches, they 
killed a buflFalo, they cooked a great deal of its flesh. 11. Then 
this way they went off, there they left him and came away, it is 
said. 12. When his war party had gone far, they changed their 
minds, crying all of them turned back. They went on and reached 
him. 13. "Now then, we have horses, we'll take you along." "No, 
even if you took me, I'd die. If you don't take me, I'll die; you 
can do nothing for me. If you walk with me, I should suffer, it 
would be bad.'^ Go away, do not turn back." 14. They left, they 
went on and on and on. They were coming, they reached Pryor, 
it is said. 15, Of these warriors, a great many were wounded, it 
is said; those not wounded were few, it is said. Those not wounded 
reached the path, those who reached Pryor were those not wounded. 
16. "Now then, go, get to camp, bring horses." They went in 
the evening. 17. Then at dawn someone began to sing praise songs, 
singing he went; this Cottontail's father was the one who sang. 
18. The large company was coming, he met this war party. 19. 
Then the man (said) : "What place exactly is he at?" "At Clark's 
Fork, the trees farthest upstream, there by the rock on the sheltered 
side among the trees. If you set out now and keep on very steadily, 
you'll get there at sunset." Then they went off, crying they pro- 
ceeded. 20. The man (said to himself): "How is he getting on? 
Is he still living perchance?" he said. Crying he went, he went 
off, it is said. He got there. They proceeded, they got to where 
he (Cottontail) was, the sun had just gone down. 21. "Have 
you still got a body, I wonder?" ^ "Come, I am not dead yet." He 
dismounted, he entered. 22. When they noticed it, his (Cotton- 
tail's) body was smaller than his leg, his leg was swollen, that 
was why. 23. There was a man named Breast, he was not a doc- 
tor, it is said, but he was experienced, it is said. With him they 
went, it is said. "Now then, for this we are bringing you, see 
what you can do about it." "That I'll do." 24. He examined it, he 
took a bullet mould. "Now, hold him in different parts" (of the 
body). They did so, it is said. He made the bullet mould pinch 
the arrowhead, he pulled, he extracted the arrowhead, it is said. 

* That is, it was swollen to an extraordinary degree. 
'' Obscure sentence. 

* That is, "Are you still alive?" 


Literature, Language, and Aesthetics 

He (Cottontail) fainted, it is said. 25. After a long time he got 
up, it is said. This man sharpened his knife, he whetted it, he 
lanced this swelling, all over he riddled it with holes, it is said, 
he pressed it. The fluid (pus) was yellow, at last it was blood, 
it is said. 26. He finished. It smarted, he (Cottontail) was out 
of his senses, towards dawn he was done. He lay sleeping, it is 
said. 27. The next morning when they observed, the next morn- 
ing there was nothing the matter with him, it is said, "c " 

(sigh of relief), he said, he snorted. He stretched himself. "I 
almost died." 28. This side of the Old Agency Mountain they 
were again in a fortified place. Again his knee was wounded, it 
is said. 29. Then near that Agency, in the area between two 
joining rivers, there they fought, again his knee was shot, it is said. 
This knee of his was not straight, it could not be bent, it was bad, 
it is said. 30. He was foolhardy, yet he was a bashful ^ young 
man, it is said, he was kind-hearted. 

31. There may have been other deeds, but we don't know about 
it exactly. 32. After a while he got a wife, she had a child. He 
divorced his wife, he was wifeless. 33. He was chasing deer, he 
fell off his horse, his knee hit the ground, his leg again swelled 
up. They made a travois and dragged it. 34. He suffered from 
the swelling of his leg. He struck his leg repeatedly. "Ghost-like 
one, you are making me suffer, (but) by your action you are not 
going to cause my death," he said, it is said. 35. Then he got 
well notwithstanding. When young men went on a raid afoot, 
when they went hunting, when they did anything whatsoever, he 
could not catch up with them. He was envious about everything; 
if he exerted himself over something, his leg would swell up again. 
It palled on him, it is said. 

36. When they got to this Agency, they were going to issue 
goods, it is said. His father had property, it is said, they made 
him distribute the goods to be issued, it is said. 37. He (Cotton- 
tail) took long strips of both red flannel cloth and black flannel. 
38. His father said: "What for, seeing that you have no wife? 
That black cloth was to be owned; you took long strips of it, you 
are doing wrong." 39. "Why, I want to get married, that is why," 
he said. The other people spoke: "He wants to marry, it's all 
right, don't talk against it." 40. Clandestinely he wanted to be- 
come a Crazy Dog, that's why he wanted to act thus. He did not 
tell, he concealed it, it is said. That's how it was. 

41. When this issuing was over, some of the camp went in that 
direction, towards Belt Hill, it is said; half the camp came down- 
stream (east), it is said. They were looking for a fight. 42. He 
wanted to come with those going east, it is said. To his sisters 

* Explained further to mean "of retiring, humble disposition." 


Literary Style of the Crow Indians 

he said: "Come, make moccasins and clothes for me. I am envious 
over everything. It is a bad business. I am bored. I'll be a Crazy 
Dog, this summer I want to die." 43. "That is something wrong 
to do. Even if you do want to die for no purpose, there are many 
enemies. If you meet some of them in the normal way and are 
not afraid, they'll kill you. If those who have become Crazy 
Dogs are not killed, they seem tmtrustworthy, they seem crazy, 
it makes them worthless," they said, it is said. 44. "Yes," he did 
not say, he did not say anything. 45. Then sometime one night 
when the camp was going to bed, he came out shouting. Then 
he sang the Crazy Dog song, it is said. 46. These sisters of his 
fell to crying, it is said. However, they could not help it. He 
sang, the whole camp had heard him. If he should quit, it would 
ruin him (i.e., his reputation). "Come, make the clothing for him." 
47. He slit a hide, he made it globular, he sewed it, he put dirt 
in; then, when the dirt was dry, he poured it out and put in 
beads, half of it he painted yellow, half he made bright red. An 
eagle tail he put on top of it (the rattle). A fox skin he made 
into a grip, he sewed little bells all over his rattle. 48. He made 
a fringe along his temples, he cut his foretop short, for his hair 
he made a false queue. He wore a very handsome necklace of 
shell, he had sea-shells for ear-rings; armlets he made for him- 
self. 49. "Half-blue" flannel was hung over his horse, he made 
a (flannel) sash for his shield, a tiger skin, and forehead discs for 
his horse. He made long metal chains. He put these on his horse, 
he made the bridle, then the reins. His horse was a bay with a 
white face, it was big. 

50. Then one evening he fitted himself out handsomely. He cut 
off his gun, making it short. He made a drum for himself, he made 
it small, he made it yellow, he slung it over his shoulder. He 
perforated the butt of his gun and made a support for it; some- 
times he put the gun into it (the support), sometimes he carried 
it next to his waist. 51. It was a fine evening, the ground looked 
yellow (with spring grass), he was again riding his horse, he came 
through camp singing, it is said. The people were most eager 
to see him. 52. He had a high-pitched ^^ voice. He was coming, 
it was some time while he was coming through camp (the Crazy 
Dogs talk crosswise), 11 they said to him: "Come that way, do 
not dance." "I certainly shall not do so," he said and reined his 
horse toward them. 53. He dismounted, it is said. He took off his 
drum, to this man he gave it, it is said. He ^~ sang. The words 
of the song were: "Come, young women, I am coming: when 

' Thus the interpreter; I should have translated "powerful." 

■ Narrator's explanatory parenthesis to clarify the following clause. 

' The musicians, not Cottontail Rabbit, as shown presently. 


Literature, Language, and Aesthetics 

the evening is pleasant, I come. Come out!" he said, it is said. He 
himself did not sing. It was those who made him dance that sang, 
thereby they made him dance. 54. Then they got to the end of 
these songs, they sang another one. At the height of their sing- 
ing 1^ he danced. He jumped around for a while, he took out his 
gun and shot at the ground, it is said. Then he shouted. He shook 
his rattle again and again. 55. He remoimted. This horse of his 
pawed the ground all over, with each hoof alternately it pawed 
the ground. When through, the horse neighed intensely. 56. The 
Crow Indians were mightily pleased. The young women liked 
him exceedingly, it is said. While the camp was moving, he would 
go singing through the camp every evening. 

57. At some time there was a very great mutual wife-stealing. ^^ 
Eats-Ears' wife was named Cherry, he (Cottontail) took her. He 
had been wifeless, now he had a wife. 58. Then some time they 
were eating, he was eating with his wife, they were eating a 
tongue. He sliced it and gave it to his wife, it is said. This young 
woman (exclaimed): "He-h6! (surprise). We are the ones who 
ought to be doing this surely; he has done it himself," she said, it 
is said. 59. "Go ahead, take it, when you mourn you will men- 
tion it," he said, it is said. This young woman took it and ate it, 
it is said. She was ashamed, it is said; however, because he had 
given it to her, she ate it, it is said. 

60. Then again some time he said to this young woman: "Why 
did you marry me? Poor dear, don't you love your forefinger ^^ 
to have married me?" he said, it is said. This young woman sat 
laughing. "That one I don't love, that's why I married you." 

61. Then on a certain evening he came singing through the 
camp, it is said. When they said, "Come this way, don't dance," 
he said, "I truly shall not do it" and dismounted, it is said. He 
(Cottontail) gave him (the musician) his drum, he took the drum 
and made him dance, it is said. 62. "Come, young women, I 
have come, when the evening is pleasant I come; come out!" He 
danced, he got to the very edge of the spectators, smiling he 
reached the end. It was when the music was at its best, it is said. 
He went shouting; shaking his rattle, he shouted. 63. He took 
his gun, he rested it on his foot, then he pulled the trigger, he shot 
at his foot, it is said. The blood was bubbling; however, he did 
not mind it, undisturbed he sang through the camp. "It bodes 
well," he said, it is said; "I was going to die." 64. Unmoccasined, 

" "When they sang best." 

" This took place early in the spring between two rival societies, a member of 
one being privileged to abduct a former sweetheart who had subsequently married 
a member of his rival society. 

^ In allusion to her having to chop oflF a finger joint when mourning his death. 


Literary Style of the Crow Indians 

his foot swelled up, he painted his foot yellow, he did not mind 
it in the least. In the evening always he would sing through the 
camp. From this foot he must have been suffering, yet he acted 
as though not noticing it. He remained unworried, it is said. 65. 
"It seems we'll never meet the enemy, already we are turning 
back," he said, it is said. "No, they are crossing ways, we shall 
meet some of them." Then they turned back at the side of Ballan- 
tine, there they camped, it is said. 

66. A man lost some horses, he turned back to where they had 
moved from, it is said. Then he saw the enemy. He came fleeing, 
it is said. He came, he reached the camp. "The enemy are over 
there!" 67. Then they (the Crow) made a sudden dash. A few of 
the Crow reached these enemies. They fought, it is said. Those 
at home, the main body of Crow also made a dash. They pro- 
ceeded and got to the enemy, it is said. They fought, it is said. 
They (the enemy) went down into a hollow, they (Crow) caused 
them to entrench themselves, 

68. For some time the Crazy Dog did not come, it is said. Then 
he came, he went on, he was close. "There he (the enemy) are 
staying entrenched," they said, it is said. 69. Cottontail Rabbit 
spoke: "That is fine, I already thought I would not see any enemies 
and that we should turn back. There I see some enemies, here is 
what I seek. When I go, strike your lips in accompaniment." When 
they proceeded (he said), "We'll dislodge them." 70. Thither 
he went, he shot into the trench, it is said. At once they hit him in 
the chest, it is said. 

71. They (Crow) took him (the corpse) back, they laid him 
down away from the river. That night it poured, these enemies 
stayed around, it is said. Their (the Crows') corpse. Young Cot- 
tontail Rabbit, lay in the water that night until daybreak, it is 
said. 72. The next morning, when they took notice, the enemy — 
however he did it — had fled and vanished, it is said. This corpse. 
Young Cottontail Rabbit, they hung over his white-faced horse, 
thus they brought him back, it is said. 73. They got to the camp 
with him, they were grieving, it is said. All the Crow, the whole 
camp, were mourning, it is said. 74. On a four-pole scaffold 
they laid him; where they laid him they tied a tipi pole to one of 
the legs, they planted it, to it they tied his sashes; his drum, his 
rattle they tied to it (the pole). Above they were blowing in the 
breeze. Then without him they moved. 

Certain features of style are inseparable from language, hence 
both versions are bound to exhibit them. Thus, indii-ect discourse 


Literature, Language, and Aesthetics 

is lacking in Crow, hence a person's speech must always be 
quoted verbatim. In other words, the vivacity conveyed in transla- 
tion would not be felt by a native audience. A noticeable 
difference falling into this category is due to a difference in 
the narrators' circumstances. Crow demands a quotative suflBx, 
-tseruk, when a statement rests on hearsay. Since No-Shinbone 
speaks as an eye-witness, this particle is not found in his version, 
whereas Yellow-Brow's bristles with it. Here, however, the matter 
is not automatically determined: in a long narrative it is not 
expected that every single sentence be followed by the su£Bx, 
hence too frequent use arouses objections, as in the case of 
another story told me by a third informant, White-Arm. My 
impression is that Yellow-Brow's style is likewise subject to 
criticism on this score; we do not encounter a single -tseruk in 
the opening part of his tale, whereas subsequently it appears 
over and over again, even in adjoining short predications. 

Insofar as the two versions cover the same ground they describe 
incidents and details very similarly. It is certainly a trivial 
difference whether the hero's rattle was made of a baking-powder 
can (No-Shinbone) or in the old style, from a piece of hide 
molded into globular shape. Both narrators describe the horse 
as bay and white-faced (No-Shinbone 13; Yellow-Brow 49) and 
mention his pawing the ground (No-Shinbone 13; Yellow-Brow 
55) as well as his metal chains. Young-Cottontail's false hair, 
necklace, and ear-rings all occur in both versions, indicating 
the importance attached to such matters by the Crow and 
their insistence on having them recounted, presumably to give 
as complete a visual picture as possible. Holman, a middle-aged 
informant to whom I read the longer variant, ridiculed the fail- 
ure to mention Cottontail's sashes before the very end. Jim 
Carpenter was not satisfied with Yellow-Brow's account either. 
I think he regarded it as too long. From our point of view the 
essential tragedy unfolded certainly seems somewhat masked 
by the wealth of detail. Both versions describe the people's ap- 
proval of the Crazy Dog and the impression made on the women; 
both mention the circumstances directly preceding his death, 
the exposure of his corpse, the burial. 

Granting that Yellow-Brow, who claimed to have been a 


Literary Style of the Crow Indians 

brother-in-law of Cottontail's (presumably in a classificatory 
sense), attempts to give the backgromid of the story, whereas 
No-Shinbone concentrates on the climax, merely referring in 
lapidary style to the cause of the hero's discontent with life, 
a characteristic difference appears when the same feature is 
mentioned. While No-Shinbone simply speaks of the horse's 
pawing the ground mightily, Yellow-Brow makes him paw with 
each hoof alternately, then neigh loudly. Judging from other 
texts secured from these informants, terseness characterizes No- 
Shinbone, while Yellow-Brow is given to epic breadth. 

Yellow-Brow's expansiveness is by no means a wholly negative 
trait. As in the example just cited, it may add to the clarity of 
the picture — explicitly stated as a desideratum by the Crow. 
The dialogue with his wife in itself adds to our conception of 
a Crazy Dog's state of mind, as does the explanation of his basic 
surfeit with life (Yellow-Brow 35) and his striking of the knee 
(Yellow-Brow 34). But this narrator undoubtedly does drag in 
material quite irrelevant to the central theme, and I find the 
threefold mishaps to the hero's knee unconvincing. Though I 
have no express statements from Jim Carpenter on the subject, 
I wonder whether his dissatisfaction rested in part on certain 
ethnographical incongruities. I should not expect an adult Crow 
to address his sisters directly (Yellow-Brow 42); and, more 
definitely, the query, why Cottontail's last wife married him is 
pointless: since he kidnapped her in the traditional fashion, she 
had no choice in the matter. Both narrators assume a certain 
familiarity with the incidents and, of course, also of Crow culture. 
Whereas Crow storytellers in general often indulge in exces- 
sive meticulousness, they also frequently leave an outsider com- 
pletely at a loss whom they are talking about. It should be noted 
that Crow normally expresses the third person simply by the lack 
of pronoun. It follows that when sentences succeed one an- 
other without nominal subjects, the auditor must have previous 
knowledge of the incidents, for he receives no aid from gram- 
mar. For example, a man extracts an arrowhead from Cottontail's 
body: "He made the bullet mould pinch the arrowhead, he 
pulled, he extracted the arrowhead, it is said. He fainted . . ." 
The last he, of course, refers not to the operator, but to Cotton- 


Literature, Language, and Aesthetics 

tail. In some instances the proper syntactic significance of a word 
appears from the context. Thus, grammatically, isbi*'9c (Yellow- 
Brow 42), "his sisters," could be the speaker of the following 
sentence, but since it appears that the speaker wants to become 
a Crazy Dog and expects his interlocutor(s) to make moccasins 
for him, "the sisters" are clearly those addressed. 

Repetition of the identical word is not regarded as a blemish 
in Crow narration; and repetition of the same thought in different 
phraseology is common, as amply illustrated by Yellow-Brow, 
here as well as in other of his texts. The latter feature is 
evidently linked with a persistent yearning for balanced structure. 
To exemplify, we find: "He divorced his wife, he was wifeless" 
(Yellow-Brow 32). "Those wounded were very many, those 
not wounded were few" {ibid. 15). "His mother died, now he 
was without mother" ( ibid. 1 ) . 

As for figures of speech, both informants use them sparingly. 
No-Shinbone compares the actions of the Crazy Dog during a 
buffalo hunt with a dog's running after cattle ( No-Shinbone 21 ) . 
He also has a hyperbolic phrase when wishing to stress the 
elaborate horse trappings, which are said to have made the horse 
invisible (No-Shinbone 9). Still more extreme is Yellow-Brow's 
statement that the swollen leg was larger than the man's body 
(Yellow-Brow 22). 

A detailed study of all my available Crow texts, many of which 
remain unpublished, would doubtless cast more light on the 
literary style of the tribe. 



Evolution and Diffusion 
in the Field of Language 

In general, nothing is easier to demonstrate than 
borrowing. Independent evolution is somewhat harder to prove, 
but in specific instances at least, it can be done. With so many 
scholars working in ethnology, it is likely that we shall eventually 
know what was borrowed from whom and by whom. It is also 
probable that in individual instances independent evolution, or 
at least an orderly series of reactions to a given set of con- 
ditions, will be adequately shown — as has already been done in 
the continuum: bride service, matrilocal residence, emphasis 
upon matrilineal descent, matriarchal clans. Similarly, in the re- 
actions of oppressed peoples towards their oppressors, as evi- 
denced by Christ himself and by many messiahs ever since. 
Or, on a humble level, by the use of animal dung for firewood 
by American Plains Indians, camel-breeders of Turkestan, and 
yak-owners in Tibet; these groups all lived where firewood was 
scarce, the weather was sometimes cold, and animal dung was 
plentiful. Various constellations of pressures do call forth identical 

This paper was found among Professor Lowie's unpublished MSS. Internal evi- 
dence suggests that it was written in the last year or two before his death. 


Literature, Language, and Aesthetics 

reactions in all kinds of peoples. What is true of culture in 
general is also true for language, since language forms a depart- 
ment of culture, though admittedly a highly specialized and in 
some respects an autonomous one. The ethnologist is therefore 
prompted to search in linguistic phenomena for parallels to the 
processes familiar to him in his wonted fields of inquiry. 

There are several relevant questions, of which the most im- 
portant are: How does evolution proceed in language — assum- 
ing that it does? Under what conditions are linguistic features 
borrowed? Who borrows from whom? Exactly what is borrowed? 
Why? What happens to the features after they are borrowed? 
What reactions are made as a result of the borrowing? Naturally, 
not all these questions can be answered for every instance that 
I shall present, but each of the cases contributes something to 
a solution of one or more problems. 

In linguistics I venture to suggest that some steps are pre- 
determined by certain preceding steps. Inasmuch as language 
grows futile when it ceases to be intelligible, whatever makes for 
intelligibility is bound to evoke compensatory developments. 
In his charming Httle book, Sound and Symbol in Chinese, the 
great Swedish sinologue, Bernhard Karlgren, tells us that in the 
sixth century Mandarin Chinese had many words that were 
distinguished in meaning from one another only by terminal 
stops. "Ka," song; "kap," frog; "kat," cut; "kak," each; were 
clearly set apart. Then, the last three words, for reasons not 
explained, lost their terminal surds, creating a number of homo- 
phones. When, on top of this, initial sonants became surds, many 
such pairs of words as "tan," arrive, and "dan," robber, were 
confounded. Furthermore, vowel sounds earlier diflFerentiated 
were likewise unified. The total result of all this leveling was that 
homophones sprang up in appalling numbers; in Mandarin the 
letter "i" had 69, and the syllable "shi" 59, meanings! Evidently, 
communication would have been frustrated but for a saving 
device — the compensatory evolution of tone. This alone did 
not bring a complete solution but required the addition of what 
Karlgren calls "special elucidative means." That is, there were 
still so many homophones that the Chinese had to use synonym 


Evolution and Diffusion in Language 

compounds — such as the pidgeon-Enghsh "look-see" — to make 
their meaning clear. 

As Karlgren points out, French oflFers a parallel in miniature 
to Chinese homophony in "cou," neck; "coup," blow; and "cous," 
[I] sew. And von Wartburg adds "ver," worm; "vert," green; 
"vers," verse; "vers," toward; "vair," squirrel fur; "verre," glass. 
Undoubtedly, Cinderella's slipper was made of "vair" (fur) and 
not "verre" (glass) as in the modern versions. Such convergence 
militates against the basic aim of speech, i.e., intelligibility, and 
when it passes a certain threshold, the development must evoke 
some defense mechanism. If the original lopping oflF of dis- 
tinguishing features results from contact with alien groups, then 
we should have an instance of a diflFusion that set in motion 
a new line of development, although perhaps nothing was ac- 
tually borrowed directly. 

A trite example that illustrates the correlation of descriptively 
disparate linguistic features is the connection of inflection with 
the position of words in the sentence. Latin requires no fixed 
word order because as a rule the nominal endings sufficiently 
distinguish the subject from the object of the sentence. "Hominem 
ursus occidit" means the same as "ursus hominem occidit"; 
neither can be translated "the man kills the bear." The roles of 
slayer and victim cannot be reversed, no matter which comes 
first. Not so, notoriously, in Chinese, French, or English, be- 
cause there are no endings to tell which is who. Sapir reminds 
us that the Chinook Indians of the Columbia River were as 
free as the ancient Romans to use any word order they liked, 
but not through an identical technique. Instead of allowing the 
nouns to establish their mutual relations, Chinook imposes a 
corresponding duty on verbal pronominal prefixes. If a language 
has no endings for its nouns, something must be done to prevent 
confusion. English and French solve the problem by using a 
fixed word order; the Chinook, by a verbal prefix. Given the 
lack of case endings, one can be certain that something will 
evolve, as an internal necessity. Evolution does, thus, take place 
in the field of language, although one could hardly predict what 
form it would take, in any given instance. 


Literature, Language, and Aesthetics 

Let us now turn to the diffusion of linguistic phenomena. 
Owing to the hmitations of my direct knowledge I shall offer 
only samples from English, German, and French. Starting from 
the point of view of the cultural anthropologist, I select from 
the preceding list of questions the following as of special im- 
portance at this point: Who borrows from whom? What is bor- 
rowed? Why? and What happens to the loan? 

The stereotyped answer to the first question has been that the 
culturally or politically superior group is the donor, and the 
culturally or politically inferior group the recipient. However, 
exceptions occur. The East African Wahima presumably Hamitic 
in origin, took over the language of the Bantu, whom they sub- 
jugated. However, by and large, the principle holds. Politically 
superior peoples do not necessarily impose their speech upon 
others, but those who think themselves inferior may voluntarily 
imitate it. Thus, as Macaulay points out, in the second half of 
the seventeenth century France achieved "an empire over man- 
kind, such as even the Roman Republic never obtained," com- 
bining "almost every species of ascendency," including linguistic 
preeminence. French became "the universal language, the lan- 
guage of fashionable society, the language of diplomacy." English 
authors "affected to use French words, when English words, 
quite as expressive and melodious, were at hand." As a par- 
ticularly odious example Macaulay quotes Dry den's lines: 

Hither in summer evenings you repair 
To taste the fraicheur of the cooler air. 

Now this dominance of French culture was even more pro- 
nounced and lasted longer in Germany than in England, be- 
cause the Germans lacked national unity and national spirit, 
such as soon asserted itself in England. The petty princes who 
aped the outer trappings of Versailles inevitably looked upon 
French literature and language as vastly superior to anything of 
native growth. Frederick the Great's partiality for French and 
his withering contempt for German literature are notorious, al- 
though he lived at a time when Goethe had begun his career and 
Lessing was closing his. Educated Germans corresponded with 
each other in French, and their own language became swamped 


Evolution and Diffusion in Language 

with French words. Since isolated cultural phenomena tend to 
persist even when out of harmony with new conditions, it is not 
surprising that the Francomania lingered long after the hegem- 
ony of the French state or of French letters. In 1847 Prince 
Metternich wrote Alexander von Humboldt in French; and it 
took Bismarck to rule that Prussian ambassadors must report to 
Berlin in German. 

The constant use of two distinct idioms is always a menace 
to the maintenance of either in complete purity. One reaction to 
this situation is the familiar mingling of words from both lan- 
guages, of which the jargon of the immigrant in the United 
States is a noteworthy example. In my youth the Germania 
Theater in New York featured a play called, "Der Corner-Grocer 
von der Avenue A," with dialogue worthy of its title. During 
the Second World War a scissors grinder, who learned of my 
Austrian birth, remarked to me, "Wir sind lucky en diesem 
Country, nicht?" In a Pennsylvania Dutch Manual — issued in 
1896 for the purpose of helping Pennsylvania-born people to 
speak English — the Ghost in Hamlet declares, "Ich bin dei 
Daddy, sei Schpook," which is neither German nor English. In 
Goethe's translation of Diderot's Le Neveu de Rameau we find 
two consecutive sentences that are literal renderings of the 
original French: "Das ist sogar unendlich wahrer, als Ihr selbst 
nicht empfindet. Ja, so seyd Ihr andern." The "nicht" has no 
place in a German sentence, and "Ihr andern" is obviously a 
rendition of "vous autres." All these instances indicate reactions 
on different levels to an intermingling of two radically different 

During the period of French ascendency, as well as earlier, 
French also borrowed words from German. The French "boule- 
vard" is the German "boUwerk"; the "lansquenet" is a "Landes- 
knecht"; a "chenapan" is a "Schnapphahn." The Swiss merce- 
naries in the French army also contributed some words, such 
as "le bivouac" (Biwache), "le cible" (Scheibe), or 'le kepi" 
(Kappi). Wartburg tells us that the Swiss soldiers were liable 
to severe fits of homesickness, hence a merryman, called a 
Bruder lustig was engaged to amuse them, whence French 
"loutic" (meaning a "wag"), already found in Voltaire. 


Literature, Language, and Aesthetics 

It would appear, then, that the question of who borrows from 
whom cannot be answered by a simple statement that the less 
potent cultural center plays a completely passive role. 

In some cases borrowing does not take place at all, when 
one would expect it to do so. Thus many American Indians 
have refused to accept English words for new ideas or new ob- 
jects but have preferred to coin their own terms by appropriate 
combinations of native stems. The Dakota word for a "horse" is 
a "mysterious dog," or a "big dog"; the Crow call a cat a "moun- 
tain-lion plus dog"; scissors are "two knives"; Sunday is "the 
day we do not work"; butter is "yellow fat"; a wagon is a "roll- 
ing object," and so on. This solution is a relatively hard one, 
and most peoples have preferred to follow the path of least 
resistance and take over the word along with the object or idea. 

Words may be accepted for any of a number of reasons, 
some of which have little connection with utility. Sometimes 
such words become incorporated into the language, and some- 
times they are soon lost. The Germans, in their wholesale borrow- 
ing of French words, took over Onkel, Tante, Cousin, and Cousine 
for uncle, aunt, male cousin, and female cousin. Naturally, the 
Germans had always had these relatives, and they had had 
names for them. These four words have, however, had a different 
subsequent history. The phonetically inadmissible Cousin dis- 
appeared altogether and has been replaced by the native Vetter. 
Cousine is still heard, but less commonly than Base, a native 
German word. Why the Germans have not reverted to Oheim 
instead of retaining Onkel, I do not know, unless it is that the 
Germanized spelling, with the elimination of the nasal, has made 
it look like a word of Teutonic origin. But there seem to me 
to have been the strongest practical reasons for clinging to Tante. 
For, strange as it may seem, German had no unambiguous word 
for an aunt. Muhme, the term now offered by purists, once 
stood for the mother's sister, while Base designated the father's 
sister. But since the sixteenth century, Base has been extended 
to either type of aunt and subsequently to female cousins, thus 
precipitating a state of confusion. The at times incredible Eduard 
Engel in 1918 lists both Base and Muhme as suitable equivalents 
of Cousine, which does not prevent him from giving Muhme 


Evolution and Diffusion in Language 

and Base as appropriate substitutes for Tante. In other words, 
no listener could know for sure which relative was meant by 
either German word. In this situation the French Tante, what- 
ever the reason for its original adoption, must have seemed a 
godsend. Similar matters of convenience have dictated the adop- 
tion of many recent loan words into German. English bond is 
used presumably because a business man cannot be expected to 
say verzinsbare Schuldverschreibung auf den Inhaber every 
time he wishes to express the idea of a bond. 

I have, then, after a somewhat sketchy fashion indicated the 
probable direction of linguistic borrowing and have discussed 
some of the motives that underlie it. Next comes the question of 
what happens to the borrowed element, in the course of its 
absorption into the language. Usually, it is first pronounced with 
only a moderately accurate imitation, then the approximate 
sound is spelled according to whatever general principles under- 
lie the native's own language, and it emerges as only a faint image 
of its former self. Thus, it is said that the strictly Viennese word 
hoppertatschet derives from de haut en bas, although by what 
processes I cannot imagine. There is little doubt that the French 
genre was converted in Austria into Schan and no doubt at 
all that potschainperl is pot de chambre. However, the Viennese 
are not the only people who thus convert foreign words into their 
own system of speech. The French turned "bowling green" into 
"boulingrin," and the English made "cherry-bang" out of "char- 

No less astounding are the transmutations in meaning. French 
vasistas — from German Was ist das? — designates a casement, sky- 
light, fan-light, shutter, or blind. Ressource in the singular usually 
denotes a social club in German. The French plumeau — feather 
duster — turns into a feather bed in Austria; bleu mourant (pale 
blue) appears in German as "bliimerant," but it means "weak" or 

One of the most puzzling things in the conversion of loan 
words is their change of gender. It is understandable that the 
English girl — used to designate a particular type in German, 
namely, a chorus girl — should be a neuter, on the analogy of 
das Mddchen or that Cafe is neuter because the German speaker 


Literature, Language, and Aesthetics 

mentally adds Haus. But why, pray, is it die Garage, when both 
Unterstellraum and Schuppen are masculine? My Duden pusil- 
lanimously gives only das Pokerspiel thus deciding the gender of 
poker by tacking a good German word onto the loan word, but 
the less timid Sprach-Brockhaus lists der Poker; perhaps this 
gender was selected in consonance with der Skat and der Tarock, 
or perhaps someone just drew it out of a hat. 

An over-indulgence in linguistic borrowing is usually the 
forerunner of reaction, suggesting again an internal and independ- 
ent evolution. I have already quoted a sample of English re- 
action to too many French words. In this paragraph I will stick 
to the reaction I know best, since it is similar to that in other 
countries and among other peoples. By the beginning of the 
seventeenth century the exaggerated use of French words in 
Germany had stimulated the beginnings of a counter-movement. 
In Weimar in 1617 a society was founded by Prince Ludwig 
von Anhalt-Kothen and three Saxon dukes for the express purpose 
of "purifying" the mother tongue. So far as I can find out, it led 
to no concrete improvements and had run its course by 1680. 
This society had successors, but none of them seem to have left 
much impression upon the language, although they doubtless 
served to keep alive the spirit of revolt. However, Joachim 
Heinrich Campe ( 1746-1818), a pedagogue and author of juvenile 
books, who in 1801 published a Worterbuch der Erkldrnng und 
Verdeutschung der Unserer Sprache aufgedrungenen fremden 
Ausdriicke, followed by a Worterbuch der Deutschen Sprache 
(1807-1811) is credited with such felicitous and current coinages 
as Zerrbild for caricature and Stelldichein for rendez-vous. With 
the rise of nationalism after the Franco-Prussian War the urge 
toward purification grew and led to the founding of the All- 
gemeine Deutsche Sprachverein in 1885, at the suggestion of 
the art historian Hermann Riegel. Among its aims it proclaimed 
the purging of unnecessary alien linguistic ingredients and the 
consequent strengthening of German unity. Besides a series of 
Germanizing glossaries for various phrases in social life, com- 
merce, or medicine, it published a journal and actually succeeded 
in influencing the vocabulary of the press and of German offi- 


Evolution and Diffusion in Language 

What struck me, as a one-time member of the organization, 
during recent trips up, down, through, and across Germany was 
the apparently complete disappearance of this relatively harm- 
less form of nationalism. It is true that a few once-common French 
words have disappeared; one no longer hears Adieu or Trottoir, 
for instance. But from Hamburg to Munich and from Bremen to 
Beyreuth there is no Haarschneider or Haarkiinstler or even 
Barhier, only Friseurs or Coiffeurs. The Speisekarte or Speisezettel 
has receded before the Menu, which bristles with such items as 
"schaschlik a la Russe," "Compote," "Pommes Dauphine," "Rump- 
steak Grandemere," and the like. When one wanted oatmeal, 
the waiters were more likely to understand if one asked for 
porridge; sl request for Haferflocken merited only a blank stare. 
In the newspapers and in everyday speech one constantly heard 
such English or American words as : best-seller, slogan ( somewhat 
camouflaged in speech behind the Teutonic pronunciation of 
"Schlo-gan"), comeback, knockout, bluff, screen star. From all of 
this I conclude that puristic fervor is a thing of the past. 

If one wishes to observe borrowing in a country where there 
seems to be no barrier against it, one should go to Switzerland. 
The German Swiss are serenely indifferent to the purity of the 
standard German that most of them can speak when they want 
to. The frequently heard "Merci vielmals" — with both words 
mispronounced — may be taken as a symbol of their attitude. 
Just as the Austrian calmly takes the technical botanical Latin 
"ribes" (for "currant"), tacks a diminutive suJBBx to it, and comes 
up with Ribisel for "grapes," so the Aleman — witness Jeremias 
Gotthelf — seizes upon trousseau and makes Trossel out of it. 

Yet the Swiss is an intransigent purist, but only as regards his 
Schwyzerdiitsch. I should perhaps explain that Schwyzerdiitsch 
is not a dialect of German, but a distinct Alamannic language, 
split into many dialects, of which Bernese is the most aberrant. 
When one wishes to speak the number "two" in Basel, one says 
zwie; in Zurich, zwo; in Bern, zwee, and in Chur, zwu, but never 
zwei. All German-Swiss learn standard German in school, and 
their newspapers are printed in it; but, irrespective of their 
educational level, they abstain completely from its use among 
themselves, except in the university lecture hall. What the Swiss 


Literature, Language, and Aesthetics 

purists fear — to judge from frequent indignant letters to editors 
— is an intrusion of standard German into their dialects, which 
they wish to keep undefiled. 

Language is a realm by itself for the linguist, who contemplates 
its phenomena rationally, but for the mere speaker it is charged 
with emotional connotations of frequently terrific power. Minor 
deviations in pronunciation from whatever is regarded as stand- 
ard precipitate ridicule, contempt, suspicion, and antipathy. 
Linguistic differences turn into cultural phenomena, into sub- 
cultural differentiae, into class badges. The Englishman who 
drops his h's is an inferior social being, and Mr. Acheson is a 
suspicious character because he speaks with a "Harvard" accent. 
Thus, language becomes, in its effect upon those who use and 
hear it, a proper subject for the ethnologist's attention. 

Moreover, one can see in the field of language the same 
phenomena that one finds in other areas of culture. Sometimes 
changes are instigated by previous developments within a lan- 
guage — as in the case of the Chinese homophones — sometimes 
they occur as a result of outside pressure — as when the Plains 
Indians invent their own words rather than borrow those of the 
white man — and sometimes they are touched off by a supposed 
insult to national pride through the adoption of too many foreign 
words — as in the German example above given. The same kind 
of development has occurred more recently in Turkey. Often 
language becomes a symbol of nationalism and passes tlirough an 
epoch of intense emotional reaction. Excesses in one direction 
lead to counter-blasts; and tlii-ough it all the man-in-the-street 
follows his naive human urge towards minimal effort and usually 
manages to offset the intentions of the intellectuals. Even when 
borrowing does not arouse resistance, it is not a routine, simple 
act. The borrowed element is gradually absorbed into its new 
linguistic background, altered in pronunciation and spelling, 
and perhaps in meaning also. Therefore, change begets more 
change, and one sees in language the same interplay of diffusion 
and evolution that occurs in any other area of human develop- 


PART III Relation of 
Ethnology to Other 

The seven papers that constitute this section can only adumbrate 
Lewie's awareness of the contribution ethnology might make to 
other fields. His interest in biology, human geography, and history 
are only glancingly revealed. His interest in psychology was real 
and persistent, but his conception of that field was essentially 
atomistic and rationalistic. Lowie saw people as "congeries of 
traits," and he distrusted the worth of the various schools of depth 
psychology that informed so much of the research in "personality 
and culture." Economics he understood largely in terms of prop- 
erty or ecology. In political science his interests focused primarily 
on social controls. "Property Rights and Coercive Powers of Plains 
Indian Military Societies" (No. 20) and "Some Aspects of Political 
Organization among the American Aborigines" (No. 21) illustrate 
this point, particularly when read in conjunction with his book 
The Origin of the State, published in 1927. These papers, like his 
much esteemed Plains Indian Age-Societies: Historical and Com- 
parative Summaries (AMNH, Anthropological Papers, XI, Part 13, 
1916), illustrate Lowie's nice appreciation for specific linkages 
between institutions and between fields of inquiry. They illustrate 
also his consummate skill as an ethnologist, in letting his data lead 
him rather than forcing them into preconceived categories. 


Psychology and Sociology 

What are the relations of psychology and sociology? 
It is clear that the sociology of both primitive and higher civiliza- 
tions yields new data for psychological interpretation. But can 
psychology as the older science, dealing v^ith more fundamental 
phenomena, throw any light on the problems that confront the 
sociologist and ethnologist? 

The question, even in this drastic form, is hardly absurd at 
the present stage of sociological and anthropological thinking. 
On the one hand, we find Graebner, the leader of the German 
historical school, resolutely turning his back on anything that 
savors of psychological interpretation. The sum and substance of 
ethnology, he tells us in his Methode der Ethnologie, is to 
determine the actual development of cultures; and this he 
forthwith outlines as the result of contact between difiFerent peo- 
ples, leading to intermixture and superposition of cultural traits. 
From this point of view any similarities observed in different 
regions must be traced to a single point of origin, for there is 
no criterion, no certain proof, of independent development, 
while cultural borrowing is not only in some cases an established 
historical fact, but may be considered demonstrated when a 

American Journal of Sociology, XXI (September, 1915), 217-229. 


Relation of Ethnology to Other Disciplines 

resemblance o£ form between the particular features compared 
is accompanied by a corresponding similarity of associated traits. 
It matters not to Graebner whether a division of society into 
exogamous moieties may mean one thing in Australia and quite 
a different thing among the Iroquois or the Tlingit. He is in- 
terested in classifying cultural results, and one moiety appears 
no different from another. If a bungling schoolboy by a double 
blunder attained the same sum as a calculating-machine, Graeb- 
ner would doubtless accuse either boy or machine of copying. 

Very different is the position assumed by such writers as 
Levy-Bruhl, Rivers, and Wissler. Each of them would insist 
that it is not a purely objective cultural epiphenomenon that 
we are dealing with in culture, but that it is precisely the sub- 
jective aspect of the problem that tempts and repays study. 
The opposition of the writers mentioned to current psychologiz- 
ing rests on very different motives and seems to be associated 
with certain notions as to the hierarchy of the sciences. Precisely 
as many biologists now hold that vital phenomena cannot be 
reduced to physics and chemistry but require a distinctively 
biological explanation, so eminent sociologists and ethnologists 
now tend to believe that sociological data are sui generis and 
defy interpretation by individual psychology. The collective 
ideas encountered by the sociologist, thinks Levy-Bruhl, are 
generically different from the ideas evolved by the individual 
mind and obey laws other than those derived from an analysis 
of individual psychology. Similarly, Wissler has suggested that 
psychological and cultural processes belong to different levels 
or cycles and should be interpreted independently of each other. 
Rivers, to be sure, does not exclude the possibility that for an 
ultimate explanation of cultural data recourse may be had to 
psychology. Nevertheless he, too, insists that in the treatment of 
immediate problems we must attempt "the correlation of social 
phenomena with other social phenomena, and the reference of 
the facts of social life to social antecedents." 

Anyone who has delved into the semi-popular ethnological 
literature of, say, the last two decades will hardly fail to sym- 
pathize in very large measure with the views just cited. The 
cheap plausibility about many current attempts to bring primitive 


Psychology and Sociology 

or modem social thought nearer to us has been admirably ex- 
posed by Levy-Bruhl. Yet the trouble with many of these in- 
terpretations is not that they are psychological but rather that 
they are folk-psychological: they rest, not on the established 
results of scientific psychology or at least on points of view 
that have proved fruitful within that science; but rather on the 
sort of oflFhand guesswork with wliich in everyday life we at- 
tempt to fathom the motives and thoughts of our neighbors. 
And even where the ethnologist does not indulge in this form 
of popular psychologizing he is likely to offer as a psychological 
explanation what cannot by the most strenuous exertion of the 
will be twisted into the semblance of one. An example is furnished 
by Professor Kroeber's "psychological explanation" of kinship 
terminologies. Professor Kroeber has it that relatives are not 
classified according to social but according to psychologico- 
linguistic categories, which he lists accordingly. His enumeration 
is one of the most notable feats in the history of the subject, 
but in what way has it anything to do with the science of 
psychology? What psychological processes cause many peoples 
to classify collateral and lineal relatives together, or to use a 
distinctive set of terms for a male and for a female speaking? 
These are linguistic phenomena that may call for a psychological 
interpretation; but merely to say that psychological factors have 
been at work is not producing the factors (such as we know 
from our textbooks on psychology), is not, then, a psychological 
explanation at all. 

Yet, when all is said and done, the spirit of skepticism that 
has invaded sociological and ethnological circles may be car- 
ried too far. I venture to believe that some facts may not only 
become more intelligible when viewed from the angle of in- 
dividual psychology, but it may be advisable not to defer this 
mode of looking at them until an indefinitely remote future. 
Even where individual psychology has not yet advanced far 
enough to give a solution of the problem, the new data may well 
prove a goad for further development of that branch of the 
science. And again an ethnologist conversant with psychology 
may give a more accurate description of his observations than 
his less sophisticated colleague. 


Relation of Ethnology to Other Disciplines 

All this becomes clearer by concrete illustration. I will begin 
by offering some remarks on a subject that figures largely in 
discussions of religion — dreams and related phenomena. Every 
sociologist knows of Tylor's attempt to account for the savage 
belief in a hereafter through visits from or to the dead as ex- 
perienced in dreams or visions. This theory can of course be 
assailed on Levy-Bruhl's principles; it may be said that an in- 
dividual will interpret his dream only in a way more or less 
predetermined by the mode of thought current in the society 
about him. But the point at issue may also be approached by 
the avenue of individual psychology. Some psychologists, such 
as Radestock and Wundt, have not hesitated to accept Tylor's 
theory at its face value. They find it perfectly natural that the 
thoughts of surviving relatives should continue to busy them- 
selves with the recently deceased, and that accordingly their 
dream life should be haunted by the figures of those who have 
just departed. Nevertheless this argument is no more than a piece 
of plausible folk-psychologizing. Yves Delage, on the basis of 
personal observations, arrived at the conclusion that ideas which 
preoccupy the mind in waking do not appear in dreams and 
that one does not dream of important events of life except when 
the period of preoccupation has ceased. More particularly he 
found that one does not tend to dream of a recently deceased 
relative.^ With qualifications that seem immaterial in this con- 
text Delage's views are corroborated by Professor Mary Whiton 
Calkins' "Statistics of Dreams." ^ She, too, finds a strong tendency 
for unimportant events of waking life to crop up in dreams, while 
events of real significance occur with amazing infrequency; and 
her independent examination of dream records confirmed Delage's 
special point as regards the apparitions of the recently deceased. 

The particular facts of this case are of course unessential. I 
have not followed recent dream-study sufiiciently to be able 
to vouch for the correctness of the views cited. But one thing is 
clear. The sociologist who is acquainted with Delage's and Profes- 
sor Calkins' observations will avoid the pitfall of a "psychological" 

^ Yves Delage, "Essai sur la theorie du reve," Revue Scientifique, XL VIII ( 1891 ), 
40 f. 

^ American Journal of Psychology, V ( 1893), 311-343. 


Psychology and Sociology 

interpretation that might otherwise seem axiomatic; the knowl- 
edge of what at least some inquirers have advanced against that 
interpretation will serve as a prophylactic against accepting 
plausible guesswork for scientific truth. 

So far, to be sure, scientific psychology carries us no farther 
than Levy-Bruhl's collective ideas. We have argued the merits 
of a particular psychological explanation from its own point of 
view and found it wanting; Levy-Bruhl's principle would pre- 
clude error by simply shutting out any explanation of this type. 
In order to vindicate the claims of scientific psychology in soci- 
ology we must therefore prove that it has something more than 
a purely corrective value. This additional function consists partly 
in the more accurate determination of facts. Ethnological and 
sociological literature fairly reeks with such phrases as "crowd 
psychology," "hypnotism," "suggestion," "influence of dream life," 
yet rarely are these terms more than exceedingly loose and 
misleading catchwords. To stick to the last-mentioned topic, in 
ordinary savage parlance such phenomena as "dreams" and 
"visions" are often thrown together under a single term. Here 
it is the duty of the field worker to discriminate lest his record 
become worthless. The difference between the religious life of 
two tribes may center precisely in the fact that in one of them 
supernatural revelations are sought through artificially induced 
visions while in the other they come through the natural medium 
of dreams. And in either case by no means all the experiences 
are of the same significance. We know that among the central 
Algonkian tribes revelations through visions may be declined 
under the influence of the preconceived notion that a particular 
kind of revelation must be secured. Here, clearly enough, a 
collective idea overrides the individual psychological experience, 
but in order fully to appreciate the significance of this very fact 
we must know definitely what the individual experience has been, 
and any analogous instances from psychology and psychiatry are 
of value. If, on the other hand, communication with the spirit 
world takes place through dreams, the question arises which 
dreams become significant, and here an intensive psychological 
analysis may become necessary. An ethnologist who knows what 
is going on in psychology may ask whether the dreams that are 


Relation of Ethnology to Other Disciplines 

culturally important conform to certain types suggested by cur- 
rent psychological discussion, say, whether they present the 
Freudian character of a repressed wish fulfilled. It may, of 
course, turn out that the dreams in question are wholly pre- 
determined by social thought; but this should be the result of 
the investigation, not a foregone conclusion. Thus, scientific 
psychology may assist in greater precision of statement as to 
recorded facts and prevent the lumping together of disparate 
phenomena; and it may further suggest lines of inquiry closed 
to those not conversant with what psychologists are doing. 

A still more important service may be rendered by psychology 
in connection with the ever-vexing problems of the unity or 
diversity of origin of similar cultural traits. Discussion of this 
point has always loomed large in the annals of anthropology; in 
recent times it has become the storm center in the whirlwind 
movements of the Graebnerian school. For, as already stated, 
Graebner denies that there are any objective criteria of in- 
dependent development. To say, for example, that the same 
mythological ideas may develop independently in different parts 
of the world seems to him worthless, subjective twaddle. The 
thing is conceivable, he admits, but this does not prove that 
it has really taken place. Now, as I have pointed out elsewhere,^ 
this is true but applies in equal measure to the supposedly ob- 
jective proof for historical connection. Here, too, what can be 
demonstrated is simply the fact that two features are similar; 
that such similarity means unity of origin is pure inference, not 
a whit less subjective than the alternative hypothesis of diversity. 
But in weigliing the evidence pro and con we cannot but attach 
great significance to whatever results scientific psychology may 
have ascertained as to general traits of the human mind. 

Take, for example, the hideous ogres that infest the mytho- 
logical world of widely separated races. Shall we accept the 
conclusion that they took shape in a single locality and thence 
spread over the entire globe? If so, how did such unrealistic 
figments of the imagination arise? And — even if we choose to 
ignore the psychology of origins — why should such fantastic 

" R. H. Lowie, "On the Principle of Convergence in Ethnology," Journal of 
American Folk-Lore, XXV (1912), 24-42. This paper is reprinted as No. 23 of the 
present work. 


Psychology and Sociology 

imagery be uncritically adopted the world over? These ques- 
tions are at least answered intelligibly, if not correctly, when we 
learn that the unrealistic figures of mythology do occur rather 
frequently in dream life. For even if their origin should not be 
traced to this source, we can at least understand why a type of 
imagery familiar from dreams should be accepted as part and 
parcel of a conceivable mythic world. This seems, indeed, to 
be the verdict of psychology. Wundt distinguishes a type of 
dreams peopled with grotesquely distorted shapes: there are 
faces with enormous probosces, projecting tongues, and gnashing 
teeth, while the head may rest dwarf-fashion on a stunted body. 
On apparitions of this type, Wundt believes, have been patterned 
the Gorgons and satyrs and pygmies of mythology.^ Other 
students support the general psychological fact. Maury often saw 
in dreams a sort of green-winged bat with a red head and a 
grimace on its face. Mourly Void reduces all these phenomena 
to a psychologico-psychological basis: in sleep tactile and motor 
sensations give rise to visual hallucinations, embracing those 
of the type now under consideration.^ 

However cautious we may be about accepting Wundt's in- 
terpretation as a final one, it is clear that the case for the 
theory of independent development becomes very much stronger 
when we find that the strange ogre figuring in myths can and 
does recur in individual dream life over and over again and may 
be referred to rather definite physiological conditions. As against 
Graebner we have thus scored a point. But the indefatigable 
disciple of Levy-Bruhl who is dogging our footsteps may object 
that when an individual dreams of, say, a Gorgon, it is because 
the Gorgon is a "collective idea" common to his social group, 
an idea with which his mind has been saturated since infancy 
and which thus naturally appears in his dreams. In other words, 
the phenomenon is essentially not psychological but sociological: 
as a modern philosopher inverted the commonplaces of ma- 
terialism by inquiring why the mind has a body, so Levy-Bruhl's 
follower nowhere sees products of individual minds becoming 
socialized but only social ideas shaping individual thought. 

Here a twofold answer is possible. In the first place, to abandon 

* W. M. Wundt, Volkerpstjchologie (Leipzig, 1900-1909), II, Part 2, pp. 114-118. 
^Nicolas Vaschide, Le Sommeil et les reves (Paris, 1911), pp. 197-225. 


Relation of Ethnology to Other Disciplines 

a psychological explanation for the Gorgon is to abandon all 
explanation; the Gorgon enthroned as a collective idea may be 
inexpugnable but it is also incomprehensible and barren. Col- 
lective ideas are blind alleys; to make headway we require the 
admittedly fragile aircraft of individual psychology. Secondly, 
Levy-Bruhl's theory involves as an essential part the doctrine 
of the diversity of human psychology. Why, then, we may reason- 
ably ask, have diverse social groups produced similar mytho- 
logical concepts? To this no answer is forthcoming from the 
opposite side. 

Let us turn now to another field of inquiry. For a long time 
ethnologists have been struck by the odd mode of associating 
ideas found among primitive peoples. In some cases, to be sure, 
the peculiarity may be due merely to our ignorance of an in- 
termediate link that has dropped out. When I am told by a 
Hidatsa Indian that the maize he plants and the wild geese 
he shoots are one and the same thing, I am puzzled; but when 
I learn that both maize and wild geese are attributes of the 
same mythic character, a logical and possible ( though not neces- 
sarily the historically correct) bond is supplied. So an educated 
Hindu might wonder at the emotional suggestions of the cross, 
but they would at once become intelligible from scriptural his- 
tory. However, in most instances the search for the missing link 
seems hopeless; and, what is more important, the very principle 
of such a search seems subject to doubt. For it assumes that 
there is a rational bond, while the trend of modern research is 
certainly to emphasize not logical but "pre-logical" associations 
and to view the rationalistic as a secondary, superimposed 

Before going farther it will be well to cite some examples of 
the types of association I have in mind. Among the Crows I 
have been told that everything in the universe exists in fours. 
As a matter of fact, the predominance of Four as the mystic 
number is very striking, not only in this tribe, but throughout an 
immense region of North America. Processions must take four 
stops; songs must be sung in sets of four; mythic heroes ac- 
complish miraculous deeds at the fourth trial, and so on. In 
Oregon, the place of Four is taken by Five, while in the Old 


Psychology and Sociology 

World both are overshadowed by Three and Seven. M. Levy- 
Bruhl has well expressed the essential fact in all these cases 
by describing the mystic numbers as categories into which reality 
is fitted: "Au lieu que le nombre depende de la pluralite reelle 
des objets pergus ou imagines, ce sont au contraire les objets 
dont la pluralite se definit en recevant sa forme d'un nombre 
mystique fixe d'avance." ® 

But it is not merely numbers that are associated with ap- 
parently fanciful ideas. To a Crow a diamond represents a 
navel cord; a rectangle, quadrilateral, right angle, and certain 
combinations of figures suggest to the Arapaho the notion of life 
and prosperity; and among the western Dakotas a form of lozenge 
symbolizes the whirlwind. 

Finally (for our present purpose), there are strange associa- 
tions with color. In addition to color associations that are self- 
explanatory, such as the connection between red and blood, 
white and snow, green and grass, there are others of a puzzling 
character. In several Plains tribes black symbolizes victory and 
joy; the Cherokees associate white with the south, red with the 
east, black with the west, and blue with the north; the Dakotas 
symbolize both the north and the south by blue, etc. 

How are we to account for such associations? The interpreta- 
tions usually given are manifestly unsatisfactory. It will not do, 
for example, to say that geometrical designs are derived from 
realistic representations of objects in nature through a process of 
degeneration, the name of the original model having been re- 
tained for the conventionalized, geometrical form. First of all, 
this does not account for symbolism of an abstract character. 
Secondly, it has been found that often the same pattern sym- 
bohzes one tiling in one tribe and another in a neighboring tribe, 
or even different things within the same tribe. As for numbers, 
it has been suggested that the mystic qualities of Four are due 
to the existence of four cardinal directions and winds, the idea 
of which is again associated with four sacred animals, colors, and 
what not. 

Levy-Bruhl has given an admirable critique of this and other 

" Lucien Levy-Bruhl, Les Fonctions mentales dans les societes injerieures ( Paris, 
1922), pp. 256 f. 


Relation of Ethnology to Other Disciplines 

so-called psychological theories of the same nature. In the 
primitive mind, he argues, there is no conception of north as a 
spatial division, with west at the left and east at the right, to 
which there are subsequently added the ideas of cold winds, 
snow, the bear, blue, etc. Rather are all these ideas bound 
up in a single complex collective idea, with the mystic elements 
masking those which we call real, and within this complex is com- 
prised the element Four. When the mystical "participations" are 
no longer felt, there are precipitated the associations that every- 
where persist in some measure. Now they are associations be- 
cause the inner bond that integrated them has disappeared; but 
originally they were of quite a different character.^ Again, the 
mystic properties of Seven among the Malays have been derived 
from the fact that the Malays believe that man has seven souls. 
Arguing in a way that must be absolutely convincing to every 
unbiased ethnological thinker, Levy-Bruhl inverts this supposed 
explanation. Seven does not play the part of the mystic number 
because the Malay believes in seven souls, but the Malay believes 
in seven souls because the preexisting numerical category pre- 
determines his speculations as to the number of souls. ^ 

Nevertheless this point of view cannot be a final one. It may be 
that the Malay conception of Seven has been an established 
category for untold aeons, and that the complex collective idea 
of Four is of corresponding antiquity in North America. Neverthe- 
less, somehow and somewhere these complex "collective ideas" 
must have taken shape in an individual mind; to "explain" them 
psychologically, i.e., to class them with related phenomena of in- 
dividual psychology, seems to be indispensable for a proper 
understanding of the facts. 

In his Inquiries into Human Faculty and Its Development, 
Francis Galton has collected data at least generically related, I 
believe, to those under discussion. He found that imaginative 
persons almost invariably think of numerals in some form of visual 

If the idea of six occurs to them, the word "six" does not sound 
in their mental ear, but the figure 6 in a written or printed form 

' Ibid., p. 242. 
« Ibid., p. 250. 


Psychology and Sociology 

rises before their mental eye. . . . Those who are able to visualize 
a numeral with a distinctness comparable to reality, and to behold 
it as if it were before their eyes, and not in some sort of dream- 
land, will define the direction in which it seems to lie, and the dis- 
tance at which it appears to be. If they were looking at a ship 
on the horizon at the moment that the figure 6 happened to present 
itself to their minds, they could say whether the image lay to the 
left or right of the ship, and whether it was above or below the 
line of the horizon; they could always point to a definite spot in 
space, and say with more or less precision that that was the direc- 
tion in which the image of the figure they were thinking of first 

To a person of this type, series of numbers arrange themselves 
"in a definite pattern that always occupies an identical position 
in his field of view with respect to the direction in which he is 
looking." These patterns or "forms" vary individually, but are 
stated in all cases to date as long back as the memory extends, 
to come into view independently of the will, and to be nearly con- 
stant for a given individual. Moreover, there is the strongest 
evidence that the peculiarity is hereditary "after allowing and 
over-allowing for all conceivable influences of education and 
family tradition." 

Galton discovered not only an association between series of 
numbers and definite patterns, but an additional association, in 
some cases, between series of numbers and colors. And what is 
perhaps of still greater immediate interest for the present purpose, 
he found that numbers are often personified and invested with 
a definite character. Three was described by different informants, 
respectively, as a treacherous sneak, a good old friend, delightful 
and amusing, etc. Galton himself "had absurdly enough fancied 
that of course the even numbers would be taken to be of the 
male sex, and was surprised to find that they were not." The as- 
sociation of color with sounds had been known prior to Galton. 
Galton notes cases of the association of definite colors with 
certain letters and with certain days of the week. One of his 
correspondents not only associated letters with colors, but con- 
versely collected "scraps of various patterns of wall paper, and 
sent them together with the word that the colour of the several 
patterns suggested to him." A blue bottle-shaped design on a 


Relation of Ethnology to Other Disciplines 

like background suggested "sweet," yellow leaves on a yellow-red 
background striated with black vertical lines meant "range." 

The psychological phenomena presented by Galton seem to 
me, I repeat, connected with the cultural phenomena under 
discussion. The association between a blue bottle design and 
sweetness does not seem to diflFer generically from the Dakota's 
association of a lozenge with the whirlwind. If an English-woman 
thinks of Tuesday in association with a gray sky color, while 
Friday suggests a dull yellow smudge, why should not the 
Indian associate the north with blue and the south with white? 
And if numbers are endowed with individual personalities by 
Europeans,^ what is marvelous in the fact that primitive tribes 
attach a preferential estimate to one (or, it may be, more than 
one) particular number? To be sure, the nature of all the as- 
sociations, individual as well as sociological, is obscure, i.e., ir- 
reducible to a logical basis. But we have at least classified the 
sociological phenomena with those phenomena of individual 
psychology that are akin to them. For that very trait emphasized 
by Levy-Bruhl as characteristic of the sociological ideas, to wit, 
their initially complex character, is in the highest degree char- 
acteristic of the Galtonian phenomena. The letter A is not first 
conceived independently by a Galtonian subject and afterward 
associated with a color. To the subject "A-brown" is an ultimate 
datum, "une representation complexe," which can indeed be 
analyzed by the psychologist, but the analysis of which can- 
not, without committing the psychologist's fallacy, be projected 
into the subject's psychological experience. 

To avoid misunderstanding, a word as to the relation of the 
psychological and sociological elements in a concrete case may 
be desirable, even at the risk of repetition. When a Crow Indian 
originating a new ceremony prescribes four sacred songs, his 
psychological condition with reference to Four may be quite 
different from that of an individual to whom Four appears as 
the incarnation of everything good and beautiful. He may be 
individually quite indifferent to the number Four; and even if 
he were not, his attitude toward it would be inextricably bound 
up with his attitude of unconsciously bowing to the traditional 

° This trait is shared by me. 


Psychology and Sociology 

category. In other words, his psychic state is characterized, in 
all probability, not by a spontaneous reaction to Four, but by 
a spontaneous reaction to the tribal lore. Substitute Three for 
Four as the tribal mystic number, and his psychic reaction 
would not vary a jot. We may go farther. Owing to the wide 
distribution of Four as the mystic number, it would be rash to 
assume that its use as such originated with the Crow Indians. 
Hence we are probably dealing, not only with the sociological 
problems of the predetermination of individual reactions by the 
social group, but also with the psychological problem of a social 
group borrowing a cultural phenomenon from another group 
which, for the sake of simplicity, we will assume to be the origi- 
nator. Now, within this hypothetical group, I repeat, the endow- 
ment of Four with certain attributes must somehow have taken 
shape in an individual mind, and the acceptance of that in- 
dividual evaluation of Four — its promotion from a psychological 
to a cultural position — is an example of the influence of the in- 
dividual on the group. That acceptance becomes the more readily 
intelligible when we recollect the highly hereditary character 
of the Galtonian phenomena and the fact that primitive com- 
munities are very largely constituted of blood relatives. 

My general conclusion as to the relation of psychology to 
sociology may therefore be summarized as follows. There can be 
no doubt that the psychological interpretation of cultural data 
is fraught with serious difficulties. We have not only to dis- 
engage the psychological fact from complicating conditions of 
a historical order, but we must also reckon with the additional 
obstacle that the individual psychic phenomenon as it confronts 
us has already been in some way molded by sociological factors. 
We may, of course, cynically eschew any and every explana- 
tion of the subjective aspect of culture. If we are not content 
to mortify the spirit to this extent, we have no choice save be- 
tween popular and scientific psychology. Scientific psychology 
will not solve all our sociological problems, nor many at the 
present time, but while not omnipotent neither is it powerless. 
It will not only act as a corrective in speculative interpretation, 
but will lend greater rigor to our formulation of fact and open 
new prospects of inquiry and explanation. 



Oral Tradition and History 


acceptance of oral traditions as historical records.^ I held then, 
as I do now, that those who attach an historical value to oral 
traditions are in the position of the circle-squarers and inventors 
of perpetual-motion machines, who are still found besieging 
the portals of learned institutions. The discussion precipitated 
by my remarks in the journal mentioned,^ and still more a great 
many private debates with fellow-students, have not shaken 
my confidence in the soundness of the views previously voiced; 
but they have shown conclusively that I had misconceived the 
psychology of the situation. Instead of being a high-priest hurling 
anathemas against the unregenerate heathen, I found myself a 
prophet preaching in the wilderness, a dangerous heretic, only 
secretly aided and abetted by such fellow-iconoclasts as Drs. 
P. E. Goddard and B. Laufer. I cannot regard it as a healthy con- 
dition of affairs in science when the adherents of antagonistic 

Address of the retiring President, delivered at the Annual Meeting of the Amer- 
ican Folk-Lore Society in New York, December 27, 1916. Printed in Journal of 
American Folk-Lore, XXX (April-June, 1917), 161-167. 

^ "Oral Tradition and History," American Anthropologist, XVII ( 1915), 596-599. 
^Ihid., pp. 599-600, 763-764. 


Oral Tradition and History 

views see no virtue whatsoever in each other's position. Per- 
chance there is some hidden source of misunderstanding that 
only need be revealed to make co-existence, if not amity, in the 
same logical universe, possible. I therefore avail myself of the 
present opportunity to present without primarily polemical intent 
the logical issues as they present themselves from my angle of 

In the first place, it may not be unnecessary to state that in 
denying to oral traditions of primitive tribes their face value, 
we are not denying to them all value whatsoever. On the con- 
trary, it is clear that even the wildest and manifestly impossible 
tales may be of the utmost importance as revelations of the 
cultural status of the people who cherish them, whether as 
annals of incidents that once occurred or as purely literary 
products of the imagination. In addition to this willingly granted 
psychological significance of such narratives, we may also admit 
a genuinely historical value, though not of the kind associated 
with this term in the present discussion. Traditions share with 
archaeological specimens, social usages, religious phenomena, 
and what not, the characteristic that likeness in distinct tribes 
calls for interpretation. Such interpretation may in many instances 
reveal beyond cavil, or at least indicate in a tentative way, an 
historical nexus otherwise unsuspected; and in such cases we are 
justified in speaking of an historical value of traditions, not in 
the sense that the traditions themselves embody truths which 
the ethnologist or folklorist must accept, but in the sense in 
which the same type of divination ritual, the same type of age- 
society, the same type of stone-axe, in different regions, may have 
an historical bearing. I will not abate one jot from this minimum 
historical estimation of tradition, nor will I concede an additional 
iota. Let us examine on what grounds such additional claims can 
be advanced. 

Against the sceptical attitude advocated by myself a very 
interesting argument has been advanced, which takes us directly 
into the heart of the problem. "Because some traditions are 
manifestly unhistorical," I have been reproached, "you rashly 
infer that no tradition has historical validity." With some claim to 
credence, I may plead that the rather elementary logical con- 


Relation of Ethnology to Other Disciplines 

siderations here advanced are not entirely beyond my ken. They 
have nothing to do with the case, however, for this rests not on 
a necessarily imperfect induction, but on more general logical, 
psychological, and methodological principles. 

That sum-total of lore which corresponds in primitive com- 
munities to what in our own culture we embrace under the head- 
ings of science and philosophy also comprises elements, in 
varying degrees of systematization, which are in native con- 
sciousness equivalent to what we call history. My general attitude 
towards these elements is simply this: If we do not accept ab- 
original pathology as contributions to our pathology, if we do 
not accept aboriginal astronomy, biology, or physics, why should 
we place primitive history alone on a quite exceptional pedestal, 
and exalt it to a rank coordinate with that of our own historical 
science? This is the, to my mind, absolutely conclusive argu- 
ment, which is independent of, though strengthened by, the 
number of cases, really tremendous, in which the glaring dis- 
parity between primitive history and our conception of the 
physical universe renders acceptance of tradition impossible. 

The really interesting problem to me is, not what degree of 
importance shall be attached to so-called historical traditions, 
but what psychological bias could conceivably make scholars 
attach greater weight to aboriginal tales of migration than to 
aboriginal beliefs as to levitation or the origin of species. While 
in the nature of the case demonstration is impossible, I have a 
very strong suspicion that lurking behind the readiness to ac- 
cept primitive for real history is the naive unconscious assump- 
tion that somehow it is no more than fair to suppose that people 
know best about themselves. This assumption, of course, need 
only be brought up into consciousness to stand revealed in its 
monstrous nakedness. The psychologist does not ask his victim 
for his reaction-time, but subjects him to experimental conditions 
that render the required determination possible. The palaeon- 
tologist does not interrogate calculating circus-horses to ascertain 
their phylogeny. How can the historian beguile himself into 
the belief that he need only question the natives of a tribe to get 
at its history? 

It may be objected that primitive astronomy and natural his- 


Oral Tradition and History 

tory do coincide in some measure with our equivalent branches 
of learning, and that consequently there is a presumption in 
favor of the view that primitive and civilized history also overlap. 
To urge this is to ignore a vital aspect of the situation. We ac- 
cept primitive observations of the stars or on the fauna or flora 
of a country as correct in so far as they conform to what we 
independently ascertain by our own methods. However, we 
neither derive the least increment of knowledge from this 
primitive science nor are we in the slightest measure strength- 
ened in our convictions by such coincidence. Exactly the same 
principle applies to the domain of history. When a Crow tells 
me that his tribe and the Hidatsa have sprung from a common 
stock, this is correct but purely superfluous information, for 
I arrive at this result with absolute certainty from a linguistic 
comparison. In history, as everywhere else, our duty is to de- 
termine the facts objectively; if primitive notions tally with ours, 
so much the better for them, not for ours. 

As a matter of fact, the case for primitive history is very 
much weaker than for primitive natural science. Natural phe- 
nomena are not only under the savage's constant observation, 
but a knowledge of them is of distinct importance to his material 
welfare. It is not strange that, say, the Plains Indians knew the 
habits of the buffalo, or should be conversant with the topography 
of their habitat. On the other hand, the facts of history are 
definitely removed from the sphere of observation when they 
have once taken place. More than that, the facts of what we 
call history are, as a rule, not facts which fall under primitive 
observation at all, but transcend it by their complexity and the 
great spans of time involved. It is as though we expected prim- 
itive man not merely to note the particular effects of rain on a 
hillside, but to form a conception of erosive processes on the 
modeling of the earth. This leads us to a point of fundamental 

There is all the difference in the world between correct state- 
ments of fact and historical truths. That my neighbor's cat had 
kittens last night may be an undeniable fact, but as a contribu- 
tion to our knowledge of present-day political and social progress 
it is a failure. That Tom Brown moved south has one meaning 


Relation of Ethnology to Other Disciplines 

when it suggests that he transferred his baggage from the 
Borough of the Bronx to a Harlem flat, and a very different one 
when the implication is that he, with thousands of his followers, 
migrated from Greenland to Patagonia. Now, my contention is 
briefly this: that the facts which we want to ascertain as his- 
torians are mainly of the latter order, while the facts recollected 
(so far as they are recollected) by primitive men are of the 
neighbor's-cat's-kittens order. In other words, I deny utterly 
that primitive man is endowed with historical sense or perspec- 
tive: the picture he is able to give of events is like the picture of 
the European war as it is mirrored in the mind of an illiterate 
peasant reduced solely to his direct observations. 

I will illustrate my contention by actual illustrations. If we 
examine an account by natives of events so recent that their 
authenticity need not be questioned, we discover what is al- 
ready known to us from other fields of inquiry; viz., that the 
aboriginal sense of values differs fundamentally from ours. 
Nothing is more erroneous than to accept uncritically, say, a 
native statement that the ceremony of a neighboring tribe is 
either akin to or different from one of his own people. A trifling 
difference in dress may lead to an assertion of complete diver- 
sity, while a superficial resemblance may lead to a far-reaching 
identification. If we glance through calendar counts and Indian 
traditions as to actual events, nothing is more striking than the 
extraordinary importance assigned to trivial incidents. Such 
things may be absolutely true, but from none of them is the 
fabric of history made. On the other hand, if we turn to oc- 
currences of tremendous cultural and historical significance, the 
natives ignore them or present us with a wholly misleading 
picture of them. Since I cannot at the present moment go 
through the entire literature of the subject, I will select a few 
instances that may fairly be taken not only as representative, 
but as constituting an argument a fortiori. 

There are few events that can be regarded as equalling in 
importance the introduction of the horse into America; moreover, 
this took place within so recent a period, that trustworthy ac- 
counts of what happened might reasonably be expected. Never- 
theless we find that the Nez Perce give a perfectly matter-of- 


Oral Tradition and History 

fact but wholly erroneous account of the case,^ while the As- 
siniboine connect the creation of the horse with a cosmogonic 
hero-myth.^ If we turn from the origin of the horse to the 
correlated phenomenon of the first appearance of the whites, 
corresponding facts stare us in the face. An Assiniboine gives 
a tale not in the least improbable of the first meeting with whites; 
only the leader of the Indians at the time is said to be the 
culture-hero.^ Among the Lemhi Shoshone I failed to find any 
recollection of Lewis and Clark's visit, but secured a purely 
mythical story about a contest between Wolf (or Coyote) as 
the father of the Indians, and Iron-Man, the father of the whites.*' 
Do we fare any better when we turn from these representatives 
of a cruder culture to peoples who have attained the highest 
status north of Mexico? Zuiii oral tradition has it that the village 
at which Niza's Negro guide Estevan lost his life, and which 
Niza himself observed from a distance, was K'iakima, In a 
masterly paper Mr. F. W. Hodge has torn into shreds the argu- 
ments advanced on behalf of the aboriginal view. He establishes 
the fact that the village in question was Hawikuh, and that 
"Zufii traditional accounts of events which occurred over three 
centuries ago are not worthy of consideration as historical or 
scientific evidence." ^ 

The general conclusion is obvious: Indian tradition is his- 
torically worthless, because the occurrences, possibly real, which 
it retains, are of no historical significance; and because it fails 
to record, or to record accurately, the most momentous hap- 

This conclusion is, I am perfectly well aware, an as yet im- 
perfect induction. To examine its ultimate validity, a special 
inquiry is necessary, for which I should like to outline the guid- 
ing principles. 

' H. J. Spinden, "Myths of the Nez Perce," Journal of American Folk-Lore, XXI 
(1908), 158. 

^ R. H. Lowie, The Assiniboine, AMNH, Anthropological Papers, IV ( 1909), 101. 

^ Ibid., Part II, p. 231. 

" R. H. Lowie, The Northern Shoshone, AMNH, Anthropological Papers, II 
(1909), 251 f. 

'' F. W. Hodge, "The First Discovered City of Cibola," American Anthropologist, 
O.S., VIII (1895), 142-152. 


Relation of Ethnology to Other Disciplines 

The historical sense of primitive peoples can be tested only 
by a scrutiny of unselected samples of their historical lore. It 
will not do, as some of our colleagues are wont, to reject mani- 
festly absurd tales and to retain those which do not contravene 
our notions of physical possibility; for by this process we get, 
in the first place, a selected series of cases, and, secondly, al- 
ready prejudge the whole matter by assuming that what is not 
ridiculously false is historically true. We must rather embrace 
in our survey every single statement which, whether miraculous 
or not from our point of view, is to the native psychology a 
matter of history. To this mass of material we must then apply 
our canons of trustworthiness; and from a comparison of the 
cases in which objective evidence supports the native statements 
with those in which such evidence is contradictory we may ar- 
rive at a statistically tenable attitude as to the general prob- 
ability of their accuracy. Had such a test been made on un- 
selected material, one of my critics would not have dared assert 
a probability of nine-tenths for native statements as to the direc- 
tion from which a tribe came. In such a test as I propose, 
aboriginal statements that a certain tribe originated in the 
very spot in which it now lives must be considered exactly on 
the same plane as any other tradition. Similarly, all statements 
of a heavenly or underground origin are of equal importance, 
for our purpose, with any other migration legends. The fact 
that they are regarded as historical by the natives is decisive 
as to their inclusion on equal terms in any such survey as I here 
suggest. Now, we know that very few of our Indians could have 
descended from the skies or climbed from an underground 
world within the period of tribal differentiation of the American 
race; and we also know that very few of them could have arisen 
in the territory they now occupy, or could have occupied it for 
very long periods. The Yuchi, for example, have no migration 
legend, and consider themselves the original inhabitants of 
eastern Georgia and South Carolina; ^ but we have recently been 
reminded that while the EngHsh colonists of 1670 refer to them 

* F. G. Speck, Ethnology of the Yuchi Indians, University of Pennsylvania, An- 
thropological Publications, I (1909), 8. 


Oral Tradition and History 

as a very powerful nation, the earlier Spanish explorers between 
1539 and 1567 mention no such tribe. ^ The assumption, con- 
sequently, is that they moved into their later habitat about the 
latter part of the sixteenth century. This case may be taken as 
typical. If events dating back tliree hundred years are no longer 
recollected, we must discount the evidence of such traditional 
lore, and cannot accept absence of migration stories as proof 
of long-continued occupancy. 

What, however, of the cases in which native traditions agree 
with objective results? The fact is simply this. The number of 
cardinal directions is four, or, if we include heaven and earth, 
six. The probability that a tribe will, in a purely mythical way, 
ascribe its origin to any particular one of these directions, is 
therefore one-fourth or one-sixth. Pending the statistical inquiries 
I have suggested, I wish to record emphatically the impression 
gained from years of experience with Indian mythology, that 
the proportion of historically correct statements wiU not be found 
to exceed that to be expected on the doctrine of chances. 

My position, then, towards oral tradition, may be summarized 
as follows: It is not based, in the first instance, on a universal 
negative unjustifiably derived from a necessarily limited number 
of instances, but on the conviction that aboriginal history is 
only a part of that hodgepodge of aboriginal lore which em- 
braces primitive theories of the universe generally, and that its 
a priori claims to greater respect on our part are nil. Such 
claims must be established empirically, if at all; but, so far as 
my experience extends, the empirical facts are diametrically 
opposed to such claims. The primitive tribes I know have no 
historical sense; and from this point of view the question whether 
they retain the memory of actual events, while interesting in 
itself, is of no moment for our present problem. The point is, 
not whether they recollect happenings, but whether they rec- 
ollect the happenings that are historically significant. Other- 
wise a perfectly true statement may be as dangerous as a wholly 
false one. If the correct description of an excursion to a northern 

' J. R. Swanton and R. B. Dixon, "Primitive American History," American Anthro- 
pologist, XVI (October-December, 1914), 383. 


Relation of Ethnology to Other Disciplines 

hunting-ground by part of a tribe is interpreted as the account 
of a permanent northern migration by the entire population, the 
result is wholly destructive of history. 

This leads us from the field of academic discussion to that of 
practical work. The question that confronts the ethnological 
practitioner is not whether primitive history in general is trust- 
worthy, but whether a particular aboriginal statement is correct 
or not. Now, what are the criteria by which its accuracy can 
be estabhshed? The only criterion that has ever been applied, 
to my knowledge, is that of physical possibility. But, as our 
Nez Perce illustration shows, this test is worthless: we simply 
shift, to use Tylor's expressive phrase, from untrue impossibilities 
to untrue possibilities. We know now that even trifling stories 
of war and quarrels are often not records of actual occurrences, 
but part and parcel of folk-lore, as their geographical distribu- 
tion clearly shows. -^^ We know the force of the human tendency 
to mingle fancy with fact, to introduce rationalistic after-thoughts, 
to ignore the essential and apotheosize the trivial, not only 
from ethnological literature, but from a study of our civilization. 
Our own historical perspective is only a slowly and painfully 
acquired product of recent years. That like other sciences it 
developed ultimately from a prescientific interest in past events, 
that in this purely genetic sense our history is an outgrowth 
of primitive tradition, is beyond doubt; but, as we cannot sub- 
stitute folk-etymology for philology, so we cannot substitute 
primitive tradition for scientific history. Our historical problems 
can be solved only by the objective methods of comparative 
ethnology, archaeology, linguistics, and physical anthropology. 

^" Franz Boas, The Eskimo of Baffin Land and Hudson Bay," AMNH Bull., XV 
(1907), 362. 



Psychology, Anthropology, and Race 

When scientists ceased to quote farmers' tales about 
the cleverness of their horses or dogs and devised laboratory 
experiments for the testing of animal behavior, a new era began 
to dawn in the history of psychology. Psychologists are laying 
aside the anecdotal method in the evaluation of individual and 
racial worth, and every anthropologist will welcome an improve- 
ment in technique that promises to shed light on one of the 
most obscure of his own problems, the question of the inter- 
relationship of empirically observed achievement and innate 
capacity. Unfortunately the psychologists who are most promi- 
nently associated with anthropological applications of their new 
tool are so ignorant of anthropology that their results are worth- 
less. It may be said on their behalf that they have been misled 
by anthropologists, that we ourselves have been guilty of spread- 
ing erroneous conceptions, but that only makes matters worse. 
The situation thus justifies an elementary consideration of the 
points at issue, a review that shall dispel the farrago of bad logic, 
bad biology, and bad faith that continues to pervade discussion of 
racial endowment. 

American Anthropologist, XXV (July-September, 1923), 291-303. 


Relation of Ethnology to Other Disciplines 

In the first place, it may be well to repudiate some absurd mis- 
conceptions, such as the strange notion that certain anthropol- 
ogists favor an extravagant influence of environmental as con- 
trasted with hereditary factors; and that they teach the absolute 
equality of all races, nay of all individuals. I do not of course 
pretend to know the views of all living anthropologists, but I am 
not acquainted with any colleague who entertains these doctrines. 
Professor Boas is commonly mentioned as the champion of such 
dogmas. When, however, I turn from the garbled account of his 
conclusions in such works as Mr. Madison Grant's The Passing 
of the Great Race to his own statements, I find nothing to support 
such misrepresentation. Professor Boas argues for "a strictly 
limited plasticity" {zugunsten einer eng begrenzten Plastizitdt) 
under the influence of an altered environment.^ On the subject 
of heredity he has this to say: 

Although we have seen that environment, particularly domestica- 
tion, has a far-reaching influence upon the bodily form of the races 
of man, these influences are of a quite secondary character when 
compared to the far-reaching influence of heredity. Even grant- 
ing the greatest possible amount of influence to environment, it is 
readily seen that all the essential traits of man are due primarily to 
heredity. ... I am inclined to believe that the influence of en- 
vironment is of such a character, that, although the same race may 
assume a different type when removed from one environment to 
another, it will revert to its old type when replaced in its old en- 

Finally, his statement as to the comparative mental make-up of 
Caucasians and Negroes is extremely cautious; he accepts the 
possibility of differences but is not convinced of such differences 
as would incapacitate the Negro for the exigencies of modern life.^ 
Personally, I take great pains to impress upon my students that 
the innate equality of all races is an unproved dogma, in spite 
of the fact that all the demonstrations of inequality hitherto at- 

^ Franz Boas, The Mind of Primitive Man (New York, 1911), p. 64; idem, Kultur 
und Rasse (Leipzig, 1911), p. 67. 

^ Idem, The Mind of Primitive Man, pp. 76 f. 
' Ibid., pp. 271 f. 


Tsychology, Anthropology, and Race 

tempted are scientifically worthless. Some time ago I formulated 
my views in the following words: 

As to the existence of superior races, I am an agnostic open to 
conviction. All evolutionists admit that at some point an organic 
change of fundamental significance occurred. It is conceivable 
that the Bushmen and Negrito, Pygmies and Negroes are organi- 
cally below the remainder of living human types, and that differ- 
ences of one sort or another even divide more closely related stocks. 
But between what is conceivable and what is definitely established 
there yawns a chasm, and where the scientist has no proof he holds 
no dogmas, though dispassionately he may frame tentative hy- 

This is not a very subtle point, but seems to transcend the com- 
prehension of some writers. One of them has even gone so far as 
to accuse me of denying innate individual differences, referring his 
readers to certain articles of mine that were expressly designed to 
illustrate these differences. 

It is an interesting fact that those who most vociferously accuse 
anthropologists of underestimating heredity as compared with 
environment are themselves the worst offenders in this regard. 
How does President Osborn, for example, account for the differ- 
ences of Cro-Magnon man in the Aurignacian and in the Mag- 
dalenian period? By the influence of environment! He writes as 
follows : 

It is probable that in the genial climate of the Riviera these men 
obtained their finest development; the country was admirably pro- 
tected from the cold winds of the north, refuges were abundant, 
and game by no means scarce to judge from the quantity of ani- 
mal bones found in the caves. ^ 

In the reduction of the stature of the woman to 5 feet 1 inch 
and of the man to 5 feet 3 inches, and in the reduction of the brain 
capacity to 1,500, we may be witnessing the result of ex- 
posure to very severe climatic conditions in a race which retained 
its fine physical and mental characteristics only under the more 
genial climatic conditions of the south. ^^ 

This is environmentalism with a vengeance! One wonders why 
those who so readily account for a difference of 300 in 

* H. F. Osborn, Men of the Old Stone Age (New York, 1915), p. 297. 
= Ibid., p. 382. 


Relation of Ethnology to Other Disciplines 

brain capacity and of iO inches in stature by a change in geo- 
graphical conditions refuse to admit that skulls may become 
somewhat narrower or wider under the influence of changed 
conditions. The difference of 10 inches in average height is about 
twice as great as the difference between the Scotch and the 
South Italians; it is greater than the difference between An- 
damanese pygmies and Frenchmen; equivalent to the difference 
between the Nilotics and the Vedda! What does Dr. Osborn 
mean? Does he believe that a climatic change effected a change 
in the germ-plasm tantamount to a heritable mutation? Or is he 
merely suggesting a "modification" in Baur's sense of the term? 
Even on the latter assumption, he is pleading for a potency of 
the environment that far transcends Boas's notion of a "strictly 
limited plasticity." 

Mr. Madison Grant is not less of an environmentalist than his 
scientific sponsor, but apparently he attributes precisely the 
opposite effects to the same climatic conditions. The Nordics, 
whom in the particular sections of the book I am now quoting 
from ^ he is pleased to favor, are said to have developed tlirough 
isolation and the selection due to the rigors of severe winters, 
wliile under "the softening influence of a life of ease and plenty" 
they succumb.^ Genial climate was necessary for the Cro- 
Magnons, the alleged spiritual forerunners of the Nordics, but 
a genial climate spells disaster for the Nordics, it seems. 

Mr. Grant, however, not merely ascribes considerable in- 
fluence to the environment when it so pleases him, but also im- 
plicitly denies the combined influence of both heredity and 
environment when the spirit so moves him. It is indeed one of 
his explicit cardinal doctrines that racial traits are "to all intents 
and purposes immutable," "fixed and rigid." He furthermore 
holds that in Sweden "there has been but a single racial type 
from the beginning" and once he even delivers himself of 
the statement that "Denmark, Norway and Sweden are purely 

" Corresponding qualifications must always be understood to accompany expo- 
sitions of Grant's views, which change from chapter to chapter, and sometimes even 
from paragraph to paragraph. 

^Madison Grant, The Passing of the Great Race (New York, 1919), pp. 38-41, 
170 f. 


Psychology, Anthropology, and Race 

Nordic." ^ Now we must recall that according to this author 
the Nordics evolved and actually flourish in the climatic con- 
ditions characteristic of their present habitation. Nevertheless he 
concludes a paragraph on the Scandinavian countries with this 
statement: "To-day all three seem to be intellectually anaemic." ^ 

As a member of the Society for the Advancement of Scan- 
dinavian Study and of the Scandinavian Club of the University 
of California, I venture to stigmatize this proposition as arrant 
nonsense. But apart from the crass ignorance it displays of the 
intellectual life of the peoples lampooned, how is such degenera- 
tion intelligible on Mr. Grant's own principles? If the Nordics are 
by heredity a favored race; if the Scandinavians are pure Nordics; 
if they "flourish, do their work and raise their families" ^^ in 
precisely the type of habitat they occupy; if racial traits "do not 
change during the lifetime of a language or an empire"; " then, 
by what magical process, neither racial nor environmental, do 
these purest Nordics degenerate to a status of intellectual anaemia 
within a few brief centuries? Perhaps Mr. Grant is not, after all, 
the champion of heredity he professes to be when it suits his 

Before leaving this writer, I will call attention to two sen- 
tences in immediate contact with each other in his chapter on 
"The Expansion of the Nordics." In the first, already quoted, 
the three Scandinavian countries are described as "purely Nor- 
dic." In the second, we are told that in southwestern Norway 
and in Denmark "there is a substantial number of short, dark 
round heads of Alpine afiinities." ^^ Comment is superfluous. 

To sum up, it is not the professional anthropologist but the 
professional heredity-monger that disregards the influence of 
heredity ad libitum. The anthropologist does not assert that 
the environment induces far-reaching effects on the germ-plasm: 
he merely asserts that certain phenomena change independently 
of the germ-plasm and in this claim he is fully supported by 

' Ibid., pp. 15, 18, 169, 211. 
" Ibid., p. 210. 
^» Ibid., p. 39. 
" Ibid., p. 15. 
^ Ibid., p. 211. 


Relation of Ethnology to Other Disciplines 

the attitude of Professor Elliot Smith, one of the few scientists 
with primarily biological orientation who have not disdained to 
try to understand the meaning of culture. ^^ 


In the past, arguments on racial diflFerences have almost always 
been advanced on the assumption that observed differences in 
cultural achievement must be the expression of correlated dif- 
ferences in inborn capacity. In one sense no one denies this; every- 
one would admit that a cat, a dog, or a monkey is incapable of 
producing or sharing in human culture. The point at issue is 
whether, when the organization adequate to the production of 
culture or, let us say, of the culture characteristic of the Upper 
Palaeolithic was reached, any further cultural advance was con- 
ditioned by equivalent changes in inborn equipment. The differ- 
ences between the material culture of, say, the West African 
Negro or the Shoshoni of Idaho on the one hand and Western 
civilization on the other are so striking that most writers naively 
assume that they are patent proofs of organic differences, and 
popular prejudice doubtless rests on the same fallacy. 

The argument is fallacious, in spite of its plausibility, for the 
following reason. When we study the known history of culture, 
we find great changes without any corresponding changes in 
racial constitution. In 1850 no one dreamt of crediting the Ger- 
mans or the Japanese people with efficiency. Elizabethan England 
was very different from the England of Queen Anne's day; and 
those who talk as though an aversion to discussions of sex were 
a deep-rooted Anglo-Saxon trait have perhaps slight acquaintance 
with Fielding and the Restoration dramatists. It is true that 
Galton asserted a racial cause for the magnificence and the 
decline of Athenian culture, but his claim is an empty allegation 
and contradictory to his own interpretation of the Renaissance. 

The instances hitherto cited involve, however, relatively slight 
differences when viewed in broadest perspective. Hence it seems 
desirable to supplement them by others. It is not merely admitted 
but contended that the Nordic race has not changed in inborn 

^^ G. Elliot Smith, "Primitive Man," Proceedings of the British Academy, VII 
( 1916), 37, 49 f. 


Psychology, Anthropology, and Race 

equipment for several thousand years except in so far as it has 
been debased by amalgamation with inferior types. Yet the 
culture of the Nordics has developed extraordinarily within the 
space of from two to three thousand years. The Cro-Magnons 
provide an even better illustration. They appeared about, say, 
25,000 B.C. and persisted through Magdalenian times, which be- 
gan about 16,000 b.c.^^ Here we have a race at least originally 
superior in inborn capacity to any now living, yet in 9,000 
years or more they cannot rise above the level of the Stone Age 
culturally! Nay, the case is still more curious, for it is the 
decadent Cro-Magnons — short and with reduced brain capacity 
— who achieve the triumphs of Palaeolithic art! 

Culture evidently does not vary with race according to any 
simple formula of functional relationship. This does not prove 
that the Tasmanians or Bushmen or Andamanese had the inborn 
capacity to develop unaided the civilization of Western Europe. 
It does prove that the difference of their culture from ours is 
not necessarily rooted in any innate difference, that the popular 
argument is wholly inconclusive. We simply do not know whether 
the evolution of Homo sapiens involved all the organic require- 
ments for any type of culture known, or whether certain de- 
ficiencies, as yet undefinable, necessarily bar certain varieties 
of the species from independently attaining such and such a 
cultural status. 

Since, then, the gross comparison of cultural achievement 
leads nowhere, so far as the determination of innate possibilities 
goes, let us turn for aid to the psychologist. Here, too, however, 
certain elementary precautions are prerequisite. 

A comparison of distinct groups involves the consideration of 
both average values and variability. It is entirely conceivable that 
two groups should coincide in their average mentality but differ 
in range, so that one may produce far more remarkable in- 
dividuals in both positive and negative direction than the other. 
Professor Fischer, for example, suggests that the Caucasian 
differs from the Negroid in precisely this point, while not excel- 

" Osborn, op. cit., pp. 18, 261, 351. 


Relation of Ethnology to Other Disciplines 

ling him in average intelligence. If this could be established, it 
would have far-reaching theoretical and practical bearings: it 
would account for the differences in cultural achievement with- 
out assuming that the average level of intelligence varies in 
different cultures; and it would imply that for the ordinary tasks 
of life the Negroid is as well fitted as the average white. 

In connection with the occurrence of extreme positive varia- 
tions it is well to bear in mind another point forcibly made 
by Father Wilhelm Schmidt, Extreme deviations from the norm 
naturally occur with greater frequency in large populations than 
in communities of several hundred. A class of fifty may have the 
average stature of the whole student body, but it is not so 
likely to have as tall members as occur in the total campus 
population of, say, ten thousand. It is not astonishing, then, 
that hordes of Andamanese or Australians numbering not over 
a hundred or two should never have produced the personalities 
which figure in the history of China, India, and Western countries. 

Another caution is of tremendous importance. Since we are 
interested in establishing the existence or non-existence of innate 
differences, the influence of training and other noncongenital 
factors, all of which for convenience' sake we may call environ- 
mental, must be eliminated. The lightheartedness, not to say un- 
scrupulousness, of many writers on this point is appalling. Ad- 
mitting, as they must, that an empirical test cannot eliminate the 
environmental factor, they decree that certain observed differ- 
ences are too great to be explained by environmental differences, 
hence are evidence of hereditary differences. The illegitimacy of 
this reasoning is apparent as soon as it is couched in clear 
language. Letting H and E represent hereditary and environ- 
mental determinants, respectively, the empirical results may be 
formulated as follows: 

H, + E, = A 
H2-\- £2 = A ± m 

It does not require a profound knowledge of mathematics to 
see that the difference ±m proves nothing as to the value of 
Hi and Ho so long as Ei and E2 differ by an unknown quantity. 
This is not academic logic-chopping pure and simple: we are 


Psychology, Anthropology, and Race 

told that Negroes are inferior to Caucasians because in certain 
tests 79 per cent of the former fell below C as against 25 per cent 
of Caucasians while only 1 per cent of the Negroes as against 
12 per cent of the Caucasians scored above C. This difference, we 
are told, is too great to be interpreted as the result of educational 
and other social differences. But New York Negroes practically 
equal Alabama whites in the tests! Hence the environmental 
factor must be taken into account, and unless we devise ac- 
curate methods for its quantitative determination, let us hold 
our tongues concerning inborn differences. 

It is a commonplace of modern science that racial and national 
groups rarely coincide. This has not deterred several prominent 
psychologists from blandly grouping immigrants into the United 
States according to their place of origin and then proclaiming 
that the results of the ensuing group tests are racial statistics. 
This is the well-nigh incredible procedure of Dr. Robert M. 
Yerkes in an article on "Testing the Human Mind," contributed 
to the Atlantic Monthly for March, 1923. Dr. Yerkes not only 
brushes aside in cavalier fashion the educational differences 
discussed in the preceding paragraphs but cites tests on Italians, 
Poles, Turks, Greeks, et al. as establishing racial differences. He 
also ingeniously suggests that the Mediterranean element ac- 
counts for the low scores of recent immigrant groups; that ele- 
ment apparently possesses the miraculous quality of detracting 
from the Italian average by its presence and from the Pohsh 
average by its absence. 

I wonder what would be thought of a naturalist who should 
wish to ascertain the characteristic weight of pure breeds of dogs 
by averaging an odd assortment of St. Bernards, dachshunds, and 
bulldogs and comparing the result with a corresponding average 
for mastiffs, fox terriers, and German police dogs. As a humble 
exercise in arithmetic the procedure may be justified, but its 
biological significance would be nil. Yet it would be better than 
Dr. Yerkes's method, for at least the naturalist would know 
precisely how many individuals of each breed he had weighed, 
but when Dr. Yerkes tests "Italians" he does not know how many 


Relation of Ethnology to Other Disciplines 

of them represent each of the relatively pure types whose inborn 
endowments he is attempting to ascertain. 

At this point I must register an emphatic protest against the 
naive assumption that because certain individuals in a region 
in which mixture of types has demonstrably occurred display 
physical features characteristic of type A they are therefore 
likewise the possessors of the mental traits that are ex hypothesi 
distinctive of the primeval "pure" type A. President Osborn goes 
further and lays down the proposition that even when one of 
the most typical traits of the Nordic, blondness, is lacking the 
individual may still be "three-fourths or seven-eighths Nordic, 
because it only requires a single dark-eyed ancestor to lend the 
dark hair and eye color to an otherwise pure Nordic strain." ^^ 
By implication dark hair and eye color will be the only features 
to dominate and the psychological traits of courage, loyalty, self- 
sacrifice and idealism innate in the Nordic will remain dominant 
in miscegenation. There is of course not a shred of evidence in 
support of such a principle of inheritance. One might well despair 
of modern biology if such slovenly pronunciamentos were not 
rejected by sane students of the subject. As Doctors East and 
Jones point out, we must be 

very cautious about drawing genetic conclusions in the human race 
based upon the possession of particular traits, in the absence of 
proof of a long-continued isolation. . . . Traits originally char- 
acteristic of certain peoples because of isolation and the consequent 
inbreeding have been shifted back and forth, combined and recom- 
bined. ... It is wholly possible, for example, that a tall, blue-eyed, 
dolichocephalic Frenchman really possesses less of the so-called 
Nordic factors than a short, dark-eyed round-head. ^^ 

Two other points may well be emphasized in this context. 
For one thing, the variability of "pure" types is largely unknown; 
we do not know, for example, how probable it is for a "pure" 
Alpine to vary so much from the norm of his type as to appear 
like a typical "pure" Nordic. Secondly, it is about time for writers 
on European anthropology to realize that things are more com- 

^ H. F. Osborn in "Preface to Second Edition" of Grant, op. cit., pp. xi f. 
'''E. M. East and D. F. Jones, Inbreeding and Outbreeding (Philadelphia and 
London, 1919), p. 250. 


Psychology, Anthropology, and Race 

plicated than a hasty perusal of Ripley's book, now twenty years 
old, may indicate. Apart from the Adriatic or Dinaric race 
recognized by many investigators, we may have other types to 
consider if Dr. Czekanowski and other anthropologists are cor- 
rect in their observations in Poland and Russia. ^^ 

Is it, then, necessary to abandon all hope of progress in this 
field? By no means: a calm survey of the difficulties merely 
leads to a formulation that does not by necessity produce absurd 
and worthless results. We cannot hope to eliminate all disturbing 
factors, but that is equally true even of such ancient sciences as 
astronomy. We can at least get rid of certain conditions that 
are bound to vitiate comparative results. 

First of all we must choose a region that is anthropologically 
well known and which has been demonstrably occupied by more 
than one racial strain, but in which strains are locally more or 
less segregated. Without assuming that it is the only country 
suitable for the purpose, I venture to suggest that Italy provides 
a very favorable starting-point. The contrast between the North 
Italian Alpine type and the South Italian Mediterranean type is 
notorious. While of course minor variations are not lacking in the 
south, the uniformity of the South Italian population is remark- 
able.^^ The hair is almost always black; the nasal index for Abruzzi, 
Campania, Pughe, and Sardinia is 69.77, 69.68, 69.49, and 68.82, 
respectively; the stature ranges provincially between the narrow 
limits of 159.9 cm. for Basilicata to 162 cm. for Campania; 
"mixed brown" pigmentation occurs in at least half of the in- 
dividuals examined, rising to 62.2 per cent in Calabria and 70.4 
per cent in Sardinia. When we consider, on the other hand, such 
typical North Italians as the Piedmontese and Venetians, we 
discover that the hair is often, if not almost always, of chestnut 
color; that the mean height is distinctly greater than among the 
Mediterraneans — 166.3 against 163.7 cm.; that there is an ap- 

" Jan Czekanowski, "Recherches anthropologiques de la Pologne," Societe d'An- 
thropologie de Paris, Bulletins et Memoires, ser. 7, Vol. I ( 1920), 48 ff. 

^* For the data that follow see V. Giuffrida-Ruggeri, "A Sketch of the Anthro- 
pology of Italy," RAI, Journal, XL VIII (1918), 80-102. 


Relation of Ethnology to Other Disciplines 

preciable percentage of individuals with fair pigmentation. In 
addition there is the marked difference in head form: the Pied- 
montese with an index of 85.7 and the Venetians with an in- 
dex of 85 are markedly brachy cephalic; the South Italians, while 
contrary to current statements not dolichocephalic at present, 
are either mesocephalic or merely of moderately brachycephalic 
character. Nevertheless, when we compare the head form of the 
several South Italian provinces, the impression of homogeneity so 
strongly suggested by other physical traits disappears; between 
the extremes represented by Sardinia with 77.5 and Campania 
with 82.1 there are intermediate figures, such as 78.4 for Calabria 
and 80.8 for Basilicata. 

These data furnish us with the possibility of sketching a program 
for psychological investigation. In the first place, it is probably 
not difficult to minimize the environmental factors: a thousand 
illiterate peasants from Sardinia will probably not differ notably 
in their cultural influences from an equal number of illiterate 
peasants from Sicily. Secondly, when we find such regional differ- 
ences in head form within an otherwise uniform population, 
they can plausibly be accounted for through racial mixture; 
specifically, the relatively broad-skulled groups are presumably 
such through the influence of Alpine mixture. The alleged innate 
mental differences are accordingly amenable to empirical verifica- 
tion or disproof: the Calabrians with an index of 78.4 may be 
assumed to be more like the Basilicatans (80.8) than like the 
people from Abruzzi (81.9) and Campania (82.1); the Sicilians 
(79.6) will be more like the Apulians (79.8) than like the other 
groups mentioned. I am well aware of the fact that very small 
differences, possibly derived from small series, may not be sig- 
nificant. It is also obvious that, with the variety of complicating 
factors, the ideal of quantitative refinement here outlined cannot 
be realized. Nevertheless, if there is anything in the alleged 
mental difference of the Alpine and Mediterranean types, the 
repeated comparison of all the otherwise homogeneous Mediter- 
ranean groups differing only by a varying degree of Alpine ad- 
mixture indicated by the cephalic index should constitute a 
crucial test. In Sardinia, with its excessively dark pigmentation, 
relatively greatest degree of dolichocephaly among the living 


Fsychology, Anthropology, and Race 

(77.5), genuine dolichocephaly (71.53) of cranial material, and 
maximum trend toward curly hair and prognathism, an especially 
favorable opportunity presents itself for ascertaining the psycho- 
logical influence of the Negroid strain that has plausibly been 
assumed as the factor determining these deviations from the 
South Italian norm. In the north, the aberrant case of Liguria, 
where the index of 79.34 stands out in marked contrast with 
that of the neighboring brachycephalic provinces, correspond- 
ing comparative tests seem desirable. 

While I have stressed the cephalic index in view of Italian 
conditions, I should not like to be interpreted as disregarding 
other physical traits. In Portugal, for example, it may well be 
that the regional distribution of blondness would provide a better 
line of cleavage than the character of the head form. 

A sane procedure will involve the systematic exploitation of 
minimal differences in conjunction with historical data. The 
Danes are known to have had largely the same antecedents as 
the other Scandinavians but they are about three centimeters 
shorter and have an index of 80.7 as against 78.5 for Norway. To 
what extent do they differ in mental make-up from other 
Scandinavians? In Norway a number of interesting problems 
arise. In sections of the country where no Lapps are known ever 
to have existed there is a marked percentage of dark-eyed 
people. ^^ This locally segregated group invites comparison with 
their typical blue-eyed "Nordic" neighbors. The latter may be 
compared with those Norwegian groups which have demonstrably 
intermarried with Lapps. Again, "pure" Lapps, such as those 
measured by Mantegazza, have an index over 87, while the 
"Lapps" of Troms, where mixture has occurred, have an index 
of 84.3, besides difi^ering in other respects. Finally, the Karelian 
Finns di£Fer appreciably from the Finns proper and might well 
be psychologically tested in comparison with them.-^ If I re- 
member Professor Retzius' statement correctly — his volume is 
not accessible to me at present — the history of the Walloons im- 
ported into Sweden is fairly well known, and certain districts 
still clearly reveal the infusion of Alpine blood. Here, then, a 

^"Halfdan Bryn, Troms Fylkes Antropologi ( Christiania, 1922), p. 19. 
^ Ibid., pp. 33, 37, 174. 


Relation of Ethnology to Other Disciplines 

comparison of Alpine and Nordic mentality may be feasible. 
No doubt many readers of this journal can suggest additional 
problems. When psychologists without bias shall have attacked 
them and arrived at statistically unexceptionable positive results, 
i.e., shall have established real innate differences, anthropologists 
will accept the conclusions regardless of their personal predilec- 
tions or prejudices. In the meantime it is their duty to denounce 
the charlatanism so prevalent in this field and to repudiate not 
biology but the sham biology that invents facts and even bio- 
logical "laws" to support personal views. 



Incorporeal Property in Primitive 

Among the problems that exercised the minds of the 
earlier evolutionists who dealt with human society, that of 
property was one of the most important. Its influence in modern 
industrial civilization was potent; hence the evolutionary sche- 
matist naturally assumed that in the earliest phases of culture 
it had been nil. Lewis H. Morgan's views may be taken as 
representative. He distinguished three major periods, — Savagery, 
Barbarism, and Civilization. The beginnings of Barbarism were 
defined by the invention of pottery, those of Civilization by the 
use of a phonetic alphabet and literary records. The two former 
periods were subdivided each into a Lower, Middle, and Upper 
Status. It was not until the Middle Status of Barbarism — ex- 
emplified by the village life of our Southwestern Indians, of 
the aboriginal Mexicans and the Peruvians — that Morgan as- 
sumed property to have played an important part. Among 
"savages," he held, property was inconsiderable. 

Their ideas concerning its value, its desirability and its inherit- 
ance were feeble. Rude weapons, fabrics, utensils, apparel, imple- 

yde Law Journal, XXXVII, No. 5 (March, 1928), 551-563. 


Relation of Ethnology to Other Disciplines 

ments of flint, stone and bone, and personal ornaments represent 
the chief items of property in savage life. A passion for its possession 
had scarcely been formed in their minds, because the thing itself 
scarcely existed. ^ 

In short, Morgan does not deny that pre-ceramic savages had 
chattels, but he minimizes the importance of the property held, 
and of the correlated acquisitive urge. His successors have gen- 
erally followed his leadership and assumed as a matter of course 
that on primitive levels property rights were weakly developed, 
there being a far-reaching communistic trend; and, specifically, 
that land was not appropriated by individuals or even familes in 
the hunting stage. 

These propositions are no longer tenable. In part they rest 
on ignorance of the ethnographic data, often on a failure to 
discriminate between moral and legal prescriptions. To illustrate 
the latter point, it is unquestionably customary to share the 
necessaries of life in a manner that sometimes amounts to prac- 
tical communism; yet, as a rule, in strict aboriginal law the line 
is clearly drawn between what is one's actual due and what is 
merely an ethical claim. There are, indeed, extreme instances. In 
northeastern Siberia a boat lying idle may be put to effective 
use without the "owner's" consent, nor is the borrower liable 
for damages in case of injury. Yet among these same popula- 
tions, other forms of property are jealously guarded from en- 
croachment. As for land, Seligmann has shown that the Vedda 
of Ceylon not only own tracts individually but practice a form of 
conveyance; and the prominence of hunting-territories among our 
Northeastern Algonkians of New England and Eastern Canada 
has been extensively described by Professor Speck.^ 

This position is fully borne out by data on two genuinely 
"savage" groups, in Morgan's sense — the Yamana (Yaghan), 
the most southerly of South American tribes, and the Semang, a 
Negrito people of the Malay Peninsula. 

The Yamana, in particular, exemplify the mingling of ethical 
and legal principles that has sometimes in the past misled 
sociological interpretation. Here there does not happen to be 

^ L. H. Morgan, Ancient Society (New York, 1877), Part IV, chap. 1. 
^ R. H. Lowie, Primitive Society (New York, 1920), chap. ix. 


Incorporeal Property in Primitive Society 

individual or family ownership of an economically valuable 
area, w^liich is held to belong to the entire territorial group. 
Certain raw materials, such as iron pyrites for fire-making and 
a species of tree whose bark was suitable for the native types 
of canoe, were restricted to definite localities; and, in these in- 
stances, utilization was permitted to territorial groups other than 
those within whose normal range these natural resources hap- 
pened to lie. Nevertheless, personal property rights were rec- 
ognized and, as usual, they rested on individual manufacture and 
effective use. Baskets must be bought from women, harpoons 
from men; as elsewhere in North and South America, even the 
children's claims to ownership are respected. While most of these 
chattels were burnt with the corpse, a dog was invariably in- 
herited by the eldest son or some other kinsman or acquaintance. 
Food is treated in the quasi-communistic fashion often reported 
for primitive tribes. In a particular case a successful seal-hunter 
immediately divided his kill into seven portions, of which he 
retained two, dividing the remainder among the five tribesmen 
present. Similarly, it is considered self-evident that the dis- 
coverer of a stranded whale should not play the part of a miser 
but should forthwith spread the glad tidings. Yet it is interesting 
to note that he had a prior claim to the booty and might select 
favored pieces or direct the distribution. However, no one was 
privileged simply to appropriate his neighbor's food, and any 
one who abused the privilege of hospitality soon fell in public 
estimation. There was also a pronounced tendency to make 
presents, whether of food, necklaces, slings, spears, or other 
implements; and acceptance involved the obligation of making 
a suitable return gift. The very fact of this institution constitutes 
proof of individual property rights.^ 

For the Semang the same general principle holds. Clothing and 
tools are personal property and can be borrowed only with the 
owner's consent. Husband and wiie pool their possessions with- 
out relinquishing their separate claims, and neither spouse in- 
herits from the other. The hut belongs to its normal builder, i.e., 
the woman, so that a divorced husband is obliged to leave it. 

' Wilhelm Koppers, Die Formen des Eigentums der Yamana auf Feuerland 
(1926), 3 Neue Ordnung, pp. 1-22. 


Relation of Ethnology to Other Disciplines 

Food is, indeed, shared with fellow-tribesmen, at least so far 
as they are related. However, two species of trees, the durian 
fruit-tree and the ipoh, which furnishes arrow-poison, are owned 
individually. To every adult male belong one or more ipohs and 
several scattered durian trees. No one would venture to tres- 
pass on these prerogatives by cutting into an ipoh or climbing 
a durian trunk. ^ 

The dogma of general primitive communism is, however, at 
once eliminated by the wide prevalence of individually owned 
forms of incorporeal property. Their very existence — sometimes 
on very rude levels, indeed, and alongside of virtual communism 
in other directions, is a noteworthy phenomenon; and the restric- 
tions on absolute ownership rights imposed by the varying mores 
of different peoples are no less interesting. 

The Eskimo may be profitably studied from this point of 
view. Like the Arctic Siberians', their hunting customs, viewed 
in isolation, might go far to support the notion that personal 
ownership is lacking. Yet the magical formulae that secure the 
Central Eskimo's luck in the chase are not shared communally. 
One man who was very successful in catching salmon stated 
that his grandmother had taught him what to sing when fishing. 
This song for salmon is also effective for seal; but for ground- 
seal he must sing another one, and still others for musk-oxen 
and for caribou. He had not taught these songs to his children, 
but intended to do so before he died. Nor were incantations con- 
fined to hunting: anciently people could use them to shorten 
their journeys, but while reciting a spell of this category they 
were not allowed to look back.^ 

Fuller data are available for the corresponding phenomena 
among the Greenland Eskimo. Spells are emphatically private 
property (en privat eg hemmelig ejendom, andre ikke maa 
bruge). They are potent not because of any spiritual agency 
but through the virtue of the words themselves, even though 
these are sometimes unintelligible. Some of the spells correspond 
to household remedies that eliminate the need for calling upon 

* Paul Schebesta, Bei den Urwaldzwergen von Malaya (1927), pp. 78 fF., 225. 
^ Franz Boas, The Eskimo of Baffin Land and Hudson Bay, AMNH Bull., XV 
(1907), 153,506. 


Incorporeal Property in Primitive Society 

a shaman to treat a patient. The largest number, however, are 
designed to secure good luck in the chase. To quote Holm: 

The charms are of great antiquity, and are as a general rule 
handed down from one generation to the other by sale. They are 
most efiFective the first time they are used and litde by little they 
lose their power; hence they must not be used except in times of 
danger, or when they are transferred to another. When the trans- 
ference takes place, none but the buyer and seller may be present, 
and in order that they may have effect, they must be paid immedi- 
ately, and dearly paid too, if there is to be any power in them; 
but then they do the possessor much benefit. The payment may 
consist, e.g., of dart points, lance points, or other costly iron work. 
As they are much reluctant to use the charms without absolute 
necessity, it is extremely difficult to get to hear them. 

One of Holm's informants recited for him a charm he had 
effectively used when on the point of death. The explorer says: 
"I paid for hearing it, otherwise it would have lost its power." ^ 

The inconsistency involved in the Eskimo position has already 
been alluded to: a communistic trend as to economic necessaries 
is coupled with strict individualism as to the magical means of 
securing food. Let us also note the limitations imposed. Effec- 
tiveness, for one thing, is contingent on purchase : in other words, 
the owner is not absolute owner in a metaphysical sense, for 
he cannot give away his spell as a gift without destroying its 
efiBcacy — a rather transparent rationalization. How far this view 
applies to the Central Eskimo instances is doubtful; possibly 
there the children have a preemptive claim to instruction in 
their elders' sacred knowledge. 

Among the Arctic Northeast Siberians similar conceptions hold 
sway. The incantations sung by the Koryak are derived ultimately 
from the Creator and have a variety of virtues, — curative, game- 
luring, and what not. They are usually in the custody of elderly 
women, who do not lightly divulge their sacred knowledge lest 
its eflficacy be destroyed. A statement of Jochelson's concerning 
the conveyance of ownership is most illuminating: "When a 
woman sells an incantation, she must promise that she gives it 

"W. C. Thalbitzer, The Ammassalik Eskimo (Copenhagen, 1914), pp. 87 ff., 
305; ibid. (1923), pp. 248-278; idem, Eskimoernes Kultiske Guddomme (Copen- 
hagen, 1926), p. 34. 


Relation of Ethnology to Other Disciplines 

up entirely, and that the buyer will become the only possessor o£ 
its mysterious power." In other words, acquisition o£ full pro- 
prietary rights involves more even than the esoteric formula; it 
requires also the transfer of a rough equivalent of "good-will." ^ 

This naturally leads to a vindication of the incorporeal char- 
acter of certain forms of property that at first blush do not ap- 
pear to merit that designation. For example, a superficial view 
of the ceremonial complexes commonly transferred among the 
Plains Indians would emphasize the material contents conveyed. 
To make the matter concrete, the Blackfoot (Montana, Alberta) 
had a series of so-called military societies, each of which was 
entered jointly by a group of approximate age-mates who pur- 
chased membership outright from an older group. Thus, the 
Hidatsa Dog organization of any period comprised individuals 
all of whom had collectively bought such badges as eagle-bone 
whistles, owl-feather headdresses, and dewclaw rattles from the 
company preceding them as owners of the ceremonial complex 
labeled "Dog." In so far as all these and other regalia were 
tangible objects, the term "incorporeal property" might be chal- 
lenged in this connection. However, closer scrutiny reveals the 
fact that a transfer implied much more than a mere purchase 
of the ordinary character. First of all, the buyers obtained the 
right to perform a specific dance, and some of their ofiicers 
gained the prerogative of appropriating any food suspended 
from the meat-racks in the camp. But even the badges themselves 
were not prized in themselves but in their proper setting: a dew- 
claw rattle, for example, could have been imitated by the Kit- 
fox or Lumpwood society, but unless duly bought in the ap- 
proved fashion from the rightful owners it was nothing but a 
travesty of the real article.^ 

This point of view appears still more clearly in the case of 
the sacrosanct complexes known as sacred bundles. The Beaver 
bundle of the Blackfoot comprises an amazing variety of dis- 
parate objects, such as skins of beavers, muskrats, and wildcats; 

' Waldemar Jochelson, Material Culture and Social Organization of the Koryak, 
AMNH, Memoirs, X ( 1908), 59. 

® R. H. Lowie, Societies of the Crow, Hidatsa and Mandan Indians, AMNH, An- 
thropological Papers, XI (1913), 225 flF. 


Incorporeal Property in Primitive Society 

skins of various birds; and so forth. Naturally, it would not be 
difficult to duplicate these elements, but it is not in them that 
the mystic potency of the Beaver bundle inheres; they are nothing 
but outward symbols of what is essential, to wit, the privilege 
to sing certain sacred songs and to perform the ritual associated 
with the objects. 

"At the formal transfer, the ritual is demonstrated as far as 
possible, four days and nights being required to complete it. 
In the normal order of events the ex-owner continues to instruct 
the purchaser for an indefinite period." ^ 

Unless a man had received this instruction he would not own 
anything genuinely valuable. In other words, he buys a series 
of prerogatives including one or more songs, the right to certain 
specific modes of behavior, knowledge of the origin myth con- 
nected with the bundle, and a tangible object or set of objects 
within a wrapping, to be guarded and opened according to 
certain rules. 

Why unsanctioned mimicry of the material parts of the bundle 
would be futile, becomes at once obvious from an exposition 
of aboriginal theory. The Blackfoot believe that every bundle 
emanates from a direct revelation by a supernatural power. 

The being appearing in the dream oflFers or consents upon re- 
quest to give power for some specific purpose. This is done with 
more or less ceremony; usually the face and hands of the recipient 
are painted, songs sung, directions given for invoking the power 
and certain obligations, or taboos, laid upon the recipient. The 
being conferring power is not content with saying that it shall be, 
but formally transfers it to the recipient with appropriate cere- 
monies. This is regarded as a compact between the recipient and 
the being then manifest, and each is expected to fulfill faithfully his 
own obligations. 

Whenever the ritual is performed, it is supposed to be a faith- 
ful replica of the initial transfer. One of the significant phenomena 
in this whole afi^air is the original visionary's right to transfer 
the contract to another, who thus acquires all his predecessor's 
rights. Only by this quasi-apostolic succession can the rapport 

' Clark Wissler, Ceremonial Bundles of the Blackfoot Indians, AMNH, Anthro- 
pological Papers, VII (1912), 100, 107, 168 ff., 272 ff. 


Relation of Ethnology to Other Disciplines 

with the supernatural world be maintained; hence an invasion 
of copyright would not help insure the blessings — longevity, 
health, and happiness — linked with authorized ownership. On 
the other hand, the genuine proprietor cannot lose the benefits 
connected with a bundle: "the bundle may be lost or destroyed 
without seriously damaging the owner, since he owns the ritual 
which is immaterial," Indeed, certain advantages cling to a former 
owner even after he has divested himself of his formal privileges. 
He may be called upon to oflBciate in a ceremony because of 
his recognized familiarity with it, or may administer a deceased 
owner's bundle until a proper transfer is consummated, and in 
either case would pocket a fee. 

As regards the transmission of sacred bundles, basic differ- 
ences divide the Plains tribes. With the Omaha and Pawnee, 
inheritance by the next of kin takes the place of the Blackfoot 
notion of transfer. While the Blackfoot do not exclude a son from 
accession to his father's ceremonial privileges, he must acquire 
them like any stranger, i.e., by the same formal acts of convey- 
ance, and these more frequently obtain between unrelated tribes- 
men, though symbolically the purchaser is regarded as the "son" 
of the seller. An interesting fusion of the two contradictory prin- 
ciples encountered in the area occurs among the Hidatsa. They, 
unlike the Blackfoot, had sons and daughters regularly acquire 
a bundle from their own father, but they must invariably pay 
him for it, one of the brothers afterwards becoming its custodian. 
In other words, children inherit the right of jointly buying pro- 
prietary rights from their own father. The latter retains the 
privilege of joining in the ritual activities, of singing the songs 
and offering prayers during a performance. Bound up with each 
bundle are a host of specific prerogatives, such as using a par- 
ticular method of painting some object in the bundle. These are 
purchased on the same occasion as the bundle, but must be paid 
for separately. "A privilege of this sort may be sold four times 
by the owner, whereupon he loses all his title to it, as among 
the Crow in corresponding cases." Here, as among the Black- 
foot, the spiritual transfer was essential. In fact, the buyer 
usually did not get the identical objective constituents of his 
father's bundle but sought to duplicate them by requisitioning 


Incorporeal Property in Primitive Society 

the services of a father's clansman; only if the latter failed in 
his quest did the father supply what was necessary. 

"It was the immaterial proprietary rights to a bundle and its 
ritual that were established by the transfer ceremony, which 
transformed a potential into an actual prerogative." ^^ 

The data for the Northern Plains tribes invite nice discus- 
sions of the basic character of "ownership" in this connection. 
On the one hand, it would appear a priori that where property 
rights are directly conferred by divine or supernatural agency 
they must be ipso facto indefeasible. That is, of course, true 
with reference to human instrumentalities. Full knowledge of the 
rituals is monopolized by the owner, and any one else speaking 
about them from observation or hearsay is not only limited in 
his information and almost bound to fall into error, but stands 
revealed as a poacher encroaching on an alien preserve. How- 
ever, the supernatural origin of the power held implied its 
revocability by the source of the blessing. Specifically, any 
infraction of the rules linked with the bundle was fraught with 
danger. For example, a Beaver bundle imposed many and onerous 
restrictions on its Blackfoot possessor. If he comes to the bank 
of a stream he must not turn back but must cross since he is 
not supposed to show fear of water in any form. Cooking must 
never be done outside his tipi, yet its sides may not be lifted, 
irrespective of the temperature. He must never blow the fire; 
in case of necessity he is allowed to blow through a pipestem. 
He cannot take back property borrowed from him. He may not 
eat of the beaver nor of the birds in the sacred pack. "The nar- 
rator," writes Wissler, "was once up in the mountains and was 
greatly famished. Finally he ventured to eat a grouse. This 
made him deathly ill." In short, the owner of sacred property 
is in many respects not its master but its slave. 

Another limitation of full ownership has already been men- 
tioned: the owner frequently is not empowered to give his 
sacred privileges away, he can only sell them, even though it 
be to his own son or daughter. Thus, the designs painted on 
members in the initiation ceremony of the Crow Indian Tobacco 

" R. H. Lowie, Sun Dance of the Shoshoni, Ute, and Hidatsa, AMNH, Anthro- 
pological Papers, XVI ( 1919), 415 ff. 


Relation of Ethnology to Other Disciplines 

society were not free for all; each method of decoration rep- 
resented a prerogative acquired in a specific vision, and was 
transferable in the usual manner of ceremonial rights. 

"GreybuU . . . had once acquired the painting privilege from 
his own mother, paying her an ermine shirt, a horse, quilts, and 
money. He sold the right to Plenty-coups for four horses." ^^ 

Perhaps still more significant is the fact that a bundle owner 
may be forced to part with his property. In some cases the obliga- 
tion is apparently moral rather than legal. Thus, Strikes-at-night, 
a Crow woman impoverished by her husband's blindness, coveted 
a Horse Dance bundle supposed to bring good luck. Direct ofiFers 
of purchase were declined by the owners. To quote my informant, 

"One owner wanted a tent and needed hides tanned. Since I 
was a good tanner, I went to his wife and offered to tan all the 
requisite hides without demanding pay outright. I got two hides 
the first time; they were large. I fixed them nicely, returned them, 
and said I would fix up the whole lodge for them." 

In this way Strikes-at-night prepared all the hides. Then the 
woman favored asked what pay the tanner wanted. "No, I want 
to take your medicine." The beneficiary said, "If you had told 
me before, I should never have let you finish the hides. Now I 
can hardly refuse you." Accordingly, she and her husband 
adopted my informant, conveying the ownership to her. "The 
other people were telling me I was very cunning because of the 
way I got the medicine." ^^ 

In the instance just cited the impression conveyed to me was 
that the compulsion, however strong in a moral sense, was not 
complete. That is to say, my informant's adopter might have 
been charged with ungraciousness had she refused to accede 
to the tanner's demands, but could not have been coerced into 
acquiescence. But among the Blackfoot the situation was differ- 
ent. A man in dire need might make a vow that if he came out 
of his difficulty safely he would buy a particular type of bundle. 

"Such appeals are usually made to the sun. The vow usually 

^ Idem, The Tobacco Society of the Crow Indians, AMNH, Anthropological 
Papers, XXI (1919), 149. 

^ Idem, Minor Ceremonies of the Crow Indians, AMNH, Anthropological Papers, 
XXI (1924), 331. 


Incorporeal Property in Primitive Society 

names a particular bundle and is registered before witnesses. 
In such cases, the owner has no option, to sell being imperative." ^^ 

What results, then, from a survey of ceremonial rights in the 
Plains is that some of them are unequivocally personal. That is 
to say, they are not shared even by the next of kin nor do they 
automatically accrue to the holder's heirs. In this respect, a 
difference obtains between them and otherwise comparable 
privileges among the Nootka of Vancouver Island. Here each 
family has its stock of songs, "no outsider being permitted to 
make use of them, unless deputed to do so by the owner. . . . 
Any woman may be hired to sing her . . . song at a menstrual 
potlatch, being paid for her services by the giver of the cere- 

While some privileges are individual in the sense that they 
might be withheld from the normal heir under unusual cir- 
cumstances, the most characteristic ones could not be diverted 
from the eldest-born son in this region of aboriginal primogeni- 
ture, so that the lineage of eldest-born descendants virtually con- 
stitutes a joint-company as regards the relevant rights. ^^ Not- 
withstanding, however, the individual nature of ceremonial 
ownership in the Plains, the religious and ethical notions bound 
up with it materially limit full property rights, sometimes even 
in a definitely legal sense. 

Equally instructive are the data from the Trobriand Islands 
off the east coast of New Guinea. ^^ In his eagerness to emphasize 
the distinctive types of ownership in different parts of the world. 
Dr. Malinowski goes so far as to regard it as "a grave error to 
use the word ownership with the very definite connotation given 
to it in our own society." Because the meaning we attach to it 
is linked with highly developed economic and legal conditions 
he infers that "therefore the term 'own' as we use it is meaning- 
less when applied to a native society." Worse than that, it 
"smuggles a number of preconceived ideas into our descrip- 

^ Wissler, op. cit., pp. 155, 174. 

" Edward Sapir, personal communication; also, idem, "A Girl's Puberty Ceremony 
among the Nootka Indians," Transactions of the Royal Society of Canada, 3d ser., 
VII (1913), 67-80. 

" Bronislaw Malinowski, Argonauts of the Western Pacific ( London, 1922 ) , pp. 
81-104, 116-120. 


Relation of Ethnology to Other Disciplines 

tion." This, however, is manifestly to exaggerate a legitimate 
point. To be forearmed against the perils of loosely applying our 
terminology, to be keenly sensible of the uniqueness of any 
particular society and its institutions, is an excellent thing. But 
we cannot coin a special word for every shade of possessory 
right as locally defined in the four quarters of the globe. It is 
far more important to define all such rights conceptually than 
to devise an infinite series of labels for them, a demand logically 
implied in Dr. Malinowski's contention, though his common 
sense prevents him from conforming to it. 

His discussion of the sociology of canoe ownership, which 
immediately follows the propositions cited, furnishes excellent 
illustrations both of his point and our qualification of it. Dr. 
Malinowski demonstrates conclusively that the toli-waga or 
"canoe-owner," to use the nearest English equivalent, is not an 
absolute owner. While he has the right to choose or eliminate 
his companions on an expedition, his maternal kinsmen "have, 
according to all native ideas of right and law, a strong claim on 
the canoe." Further, even unrelated patricians of the community 
could not easily be excluded in the absence of special cause; 
and still others would have a moral "de facto right to sail" be- 
cause of their skill as mariners. Again, it is the toli-waga that 
assembles the council and broaches the question as to the date 
of sailing. However, this right of initiative, on closer scrutiny, 
turns out to be purely nominal, since "both in construction and 
sailing, the date of enterprise is determined by outward causes, 
such as reciprocity to overseas tribes, seasons, customs, etc." 
This same sort of relative property right is distinctive of the 
"Kula," an extraordinary system of exchange by which arm- 
shells and necklaces ceremonially prized are exchanged, each 
gift being repaid by an equivalent counter-gift. Here, Dr. Mali- 
nowski points out that the recipient never retains his acquisi- 
tion for any length of time, — never for more than a year or 
two, and even this moderate period "exposes him to the reproach 
of being niggardly." In other words, public sentiment demands 
that valuables of this type be kept in circulation. Evidently our 
authority is warranted in saying that this sort of "ownership," 
esteemed because of the renown coupled with even fleeting 
possession, is sui generis. 


Incorporeal Property in Primitive Society 

Notwithstanding this admission, however, there is evidence 
that ownership quite as complete as any found in any com- 
munity coexists with those more Hmited forms that have so 
deeply impressed their reporter. Let us return to the toli-waga. 
However he may be restricted in the practical utilization of a 
boat, the honorific title is indefeasibly his. Even when his 
closest maternal kinsmen collectively apply it to themselves, 
"this would be an abuse of the term." Further, "the mere privilege 
of using exclusively this title is very highly valued by the natives"; 
and though the right of summoning the council and inaugurating 
a voyage is admittedly nominal, "the formal privilege is strictly 
confined to the toli-waga, and highly valued," quite apart evi- 
dently from the appreciable economic perquisites of the office. 

To turn to other phases of culture. In the Trobriands, myths 
are not owned quite so exclusively as in certain other areas, yet 
particular ones are associated with lineages who "are supposed 
to possess the most intimate knowledge of the mythical events, 
and to be an authority in interpreting them." Dances are more 
definitely individual property, the original inventor having the 
right to perform it in his village. "If another village takes a 
fancy to this song and dance, it has to purchase the right to 
perform it." Similarly, magical power — the knowledge of formu- 
lae intrinsically potent to achieve desired ends — is rated as a 
form of property. A very interesting analogy (though with a 
difiFerence) to certain Plains Indian conditions may be noted. 
Sometimes matrilineal blood-relatives, who would be the natural 
heirs under aboriginal law, desire to secure certain goods in their 
elder's lifetime. In such cases substantial payments must be made 
by the nephew or the younger brother, e.g., the magic may be 
taught bit by bit in return for payment in instalments. "After 
the final payment, the title of ownership is definitely handed over 
to the younger man." '^^ The difference from the Hidatsa bundle 
concept lies in the fact that apparently material as well as in- 
corporeal goods may be thus acquired by a Trobriander; and, 
further, an Hidatsa could not inherit the bundle except by mak- 
ing the customary payment. 

Disabilities on sex lines introduce us to another category of 
incorporeal privileges. In the Banks Islands of Melanesia, women 

" Ibid., pp. 185 fF., 291, 317, 329. 


Relation of Ethnology to Other Disciplines 

never drink or prepare kava, and only in quite recent times have 
they been allowed to watch its preparation. "I never saw my 
informant more heartily amused," writes Rivers, "than when I 
told him that I had seen kava being made by a woman in 
Samoa." ^^ This evidently must be taken in conjunction with the 
widespread Melanesian and Australian custom of eliminating 
women from ceremonial life, for on the Banks Islands the use 
of kava was formerly restricted to those of high rank in the 
men's fraternity. The sexual division of labor, involving as it 
does an allotment of onerous tasks, seems to have little in common 
with the notion of property rights as commonly understood. 
But it may also imply greater or lesser kudos, and frequently 
very practical prerogatives. The Formosan women raise millet 
and sweet potatoes, and it is they, not their husbands and 
brothers, who superintend the granaries, dealing out daily sup- 
plies to the female representatives of the several households. ^^ 
On the other hand, in many African and Siberian tribes man's 
concern with the care of cattle disqualifies women from owner- 
ship, and cases are known in which remote kinsmen take pre- 
cedence of daughters in the inheritance of livestock. In short, 
sex often includes the right to potential ownership of goods, 
corporeal or incorporeal. 

This is naturally only a special form of group ownership. 
Wherever succession to dignity is regulated by some such prin- 
ciple as primogeniture, the entire group of incumbents, actual 
and potential, may be conceived as a corporation, the entire 
membership sharing the same privileges, even though possession 
and usufruct be limited to one individual at a time. It is this 
sense of a group interest that of course tends to nullify testa- 
mentary dispositions purporting to override established precedent. 

A quaint coupling of prerogative with disability is reported 
from a people of the Upper Nile region. Though a Lango woman 
is never allowed to hold any but personal belongings, she may 
veto her husband's proposal to give away a single head of stock, 
provided it was obtained as a part of her daughter's bride-price. 

" W. H. R. Rivers, The History of Melanesian Society (Cambridge, 1915), p. 82. 
"J. B. M. McGovern, Among the Head-Hunters of Formosa (London, 1922), 
pp. 124 ff. 


Incorporeal Property in Primitive Society 

The purpose of the custom is to safeguard the matrimonial 
prospects of a woman's son, for here, as elsewhere in Africa, the 
payment for a daughter is supposed to provide her brother with 
a wife. To put it diflFerently, a woman who has given birth to a 
girl and married her off acquires thereby a limited control over 
the compensation offered.^'' 

The Lango furnish us with another instructive sample of in- 
corporeal property in disguise. Hunting-territories are owned 
individually, and without the owner's consent encroachment by 
other hunters would be illegal. However, there are responsibilities 
acccompanying prerogative. The tract must be surrounded by a 
fire-break to ward off conflagrations spreading to or from the 
plots of fellow-tribesmen. What, it may be asked, is there in- 
corporeal about land? The point is that no man really owns "his" 
land. Anyone desiring to build on it or to reserve a plot thereon 
for tillage cannot be denied, even though formal permission 
must first be sought. In other words, "the won arum owns the 
hunting rights over the land rather than the land itself." How- 
ever, he is not liable for damage to the newcomer's house or 
crops .^° This instance recalls comparable data from New Zealand, 
where the same territory was differently exploited by different 
households. One family would have a monopoly of the claim 
to the shellfish or berries found, another reserved the right to dig 
fern-roots or to hunt rats.^^ 

Enough has been said to demonstrate the reality of incorporeal 
ownership on the level of illiterate peoples. How vital a part it 
plays in their lives is at once apparent when we recall the 
manifold ramifications of the subject inevitable even in tliis 
brief exposition. Starting from a juridical concept, we have had 
to touch the entire scope of cultural phenomena, — not only the 
proximate fields of social structure and government but the more 
remote departments of economics, industry, arts, and religion. 

'J. H. Driberg, The Lango, a Nilotic Tribe of Uganda (London, 1923), p. 172. 
Ibid., pp. 112, 171. 
Lowie, Primitive Society, p. 229. 



Economic Factors and Culture 

Utilitarian motives loom so large in contemporary 
life as to veil the force of other psychological drives. Thus an 
economic determinism has found favor in many circles far beyond 
the bounds of Marxian philosophy. As usual, sanity lies neither 
in spurning explanations that have appealed to serious scholars 
nor in clasping them fervently to our bosom as the long hoped 
for key to all mysteries, but in a discriminating appraisal of what 
is really solved and what eludes solution. 

That man must have food in order to enjoy any social life 
whatsoever is an unchallenged truism. Yet even this can be 
invested with meaning by demonstrating the concrete way in 
which it affects particular situations. Among the lower hunters 
who are constantly obliged to shift their camps from the sheer 
need for food there are nevertheless periodic major festivals. 
How, on this level, is it possible to unite even a hundred people 
in one spot for a number of weeks or months? Evidently only if 
some lucky fluke creates a temporary surplus of food. In Tierra 
del Fuego, accordingly, it is the unpredictable stranding of a 
whale that precipitates a large assemblage and with it the pos- 

This paper was found among Professor Lowie's unpublished MSS. Internal evi- 
dence suggests that it was written in 1939 or 1940. 


Economic Factors and Culture 

sibility of an initiation festival.^ Again, among the Murngin of 
Northern Austraha the great ceremonies never take place during 
the rainy season, nor at the beginning of the dry season when 
the grass is too high for travel and the food plants are not yet 
ready for harvesting. It is necessary to wait until the grass can 
be burned and the women can collect wild yams, lily bulbs, and 
cycad nuts as provisions for the ceremonial period.^ Similarly, 
festive activities of the Shoshoneans of Nevada were restricted 
to occasions when "food supplies were sufficient to support an 
abnormally large number of persons for a week or so." This con- 
dition could be created by communal antelope hunts, rabbit 
drives, or pine-nut gathering.^ 

So far, so good. Ceremonialism would not be possible for our 
three tribes without the guarantee of ample food. But what of 
the ceremonial itself? Its alleged ends are partly economic: 
Australians may attempt to multiply game animals, Shoshoneans 
to increase the salmon run. But the ritualistic activity in its 
totality remains untouched by the economic theory. Why are 
women frightened away from the dance ground by the hum of 
bull-roarers? Why is the emphasis on the initiation of boys when 
so many American tribes celebrate only the girls' coming of 
age? The terrorizing of Australian women might be made to 
fit into the Marxian pattern as the exploitation of a downtrodden 
sex duped into furnishing provisions for the male celebrants; 
but why are they similarly cowed by the Ona of Tierra del 
Fuego, where practically all food is secured by men? Why do 
the Ona lavish infinite pains on the elaboration of masquerade 
costumes for their festival? These are but a few of myriad ques- 
tions that arise: they and others like them form the core of the 
problems that arise for scholars concerned with the study of 
ceremonialism. They are not answered by the axiom that the 
phenomena would be precluded if the would-be-performers 
were dying of starvation. Nor is it any help to be told that the 

^ Martin Gusinde, Die Yamana; vom Leben und Denken der Wassernomaden am 
Kap Hoorn (Modling bei Wien, 1937), pp. 823 f. 

^William Lloyd Warner, A Black Civilization (New York and London, 1937), 
pp. 340, 347. 

*J. H. Steward, Basin-Plateau Aboriginal Sociopolitical Groups, BAE, Bull. 120 
( 1938), pp. 45 f., 237. 


Relation of Ethnology to Other Disciplines 

celebrants are seeking satisfaction of vital needs. No one denies 
that their activities are motivated: the problem is whether the 
needs and motives are preponderantly, nay, exclusively, bound 
up with the food-quest; and this is clearly contrary to all ex- 

One main objection to the economic interpretations of culture 
that have hitherto been oflFered is that they fail to come to grips 
with those problems that obtrude themselves spontaneously 
on any unprejudiced observer of the data. The theorists of this 
school ignore even so obvious a fact as the irrational ingredients 
of economic activity, a fact so convincingly established by the 
late Eduard Hahn and so overwhelmingly corroborated by later 
inquiry. To take a single example, East Africans are enthusiastic 
stock-breeders; but is their animal husbandry to be gauged by 
our standards? Far from it. A Shilluk keeps hundreds of cattle, 
yet slaughters them so rarely that he is obliged to maintain his 
hunting tecliniques for an adequate supply of meat. His small 
cows yield but little milk, his oxen normally serve no purpose 
at all. But these Negroes, who have failed to perfect their dairying 
industry and who eschew a beef diet, expend enormous effort on 
massaging the humps of their beasts and twisting their horns 
into grotesque shapes.* 

In other words, man spontaneously exerts himself not merely 
in order to fill his belly, but also for utterly fanciful objectives. 
If the domain of economic life itself is shot through with non- 
economic motivation, it is a fortiori hopeless to derive art, litera- 
ture, philosophy, and religion from an exclusive interest in food 
and profit. 

Nothing is more certain to prejudice a sound economic inter- 
pretation than such unsupportable claims on the part of enthusi- 
asts like Dr. Paul Radin, whose merits as a field investigator must 
not blind us to theoretical biases.^ The remarkable thing about 
this attempt to account for aboriginal beliefs is its persistent 
neglect of economic facts while rendering on every other page 
lip service to economic determinism. The picture of the Central 

* Wilhelm Hofmayr, Die Schilluk ( Modling bei Wien, 1925 ), passim. 
^ Paul Radin, Primitive Religion; Its Nature and Origin (New York, 1937), esp. 
pp. 51 S. 


Economic Factors and Culture 

Eskimo is typically Voltairean: His shamans, organized to cow 
the laity, terrorize the masses by an elaborate mechanism of be- 
liefs, formulae and taboos. They have deliberately devised the 
system to keep all contact v^ith the supernatural in their own 
hands and to exploit the superstitious fears of their dupes. Thus, 
the shaman may abduct his neighbors' wives and tyrannize them 
generally with impunity since no one dare rebel against, let alone 
murder, a communicant with the spirit world; accordingly, he 
conquers economic insecurity and "life flows on for him in com- 
parative ease." 

Not a single one of these propositions bears even a remote re- 
semblance to the facts. The Eskimo shamans are not organized; 
they may bully their communities, but a medicine-man suspected 
of kidnapping souls is killed; and ordinary people possess talis- 
mans and magical formulae, so that supernatural power is by no 
means confined to the shamans. An old Iglulik woman who had 
inherited a famous magical spell was able to achieve "economic 
security" by selling it to a hunter, who pledged himself to provide 
her with food and clothing as long as she lived. In other words, 
she got herself an annuity.*' 

The basic error, however, lies deeper. Dr. Radin wholly disre- 
gards the vital fact that the insecurity of which he speaks does 
not exist in a social sense among the Eskimo: so long as food is 
available no one suffers from want of the bare necessities. Even 
the idler gets his meals; he simply sinks in public esteem. That is 
the typical primitive attitude, as amply attested for Melanesia.^ 

In short, the "wealth" of an Eskimo shaman or a Melanesian 
magician does not make him more secure as to the bare neces- ^ 
sities of existence; it simply confers prestige. But prestige is an / 
ideological phenomenon. -- ^^-'^ ■- 

The correct approach to an economic interpretation is not to 
deal in vague generalities, but to isolate specific elements in the 
productive and distributional system of peoples and to correlate 

* Franz Boas, The Eskimo of Baffin Land and Hudson Bay, AMNH Bull., XV 
( 1907), 117 f.; Knud Rasmussen, Intellectual Culture of the Iglulik Eskimo (Copen- 
hagen, 1929), pp. 144, 149, 165. 

'Rasmussen, op. cit., p. 159; Kaj Birket-Smith, The Caribou Eskimos (Copen- 
hagen, 1929), p. 261; Hortense Powdermaker, Life in Lesu (New York, 1933), 
p. 224. 


Relation of Ethnology to Other Disciplines 

these with eflFects that can be reasonably traced to them. In this 
inquiry a much finer analysis is required than is customary. Thus, 
the influence of "agriculture" is not at all what it is commonly 
supposed tp be. In its incipient phases "agriculture" produces no 
spectacular results either in density of population, or stability, 
or cultural superstructure. Because aboriginal farmers cannot 
work heavy soils with crude dibbles, they avoid rich lands in 
favor of inferior ones. In consequence, less than one per cent of 
the possible area within the Eastern United States was cultivated 
by the indigenous redskins; and their density of population fell 
below that of purely hunting Californians.^ 

That "farming" in itself harbors no magical potency is likewise 
indicated by Lapp experience. The sedentary Lapps of Kauto- 
keino are decidedly below par as compared with their nomadic 
brethren. One of these farmers raises not cereals for human con- 
sumption, but fodder; his livestock comprises typically a few 
sheep, a horse, from two to a dozen cows; he produces little or 
no butter and cheese. Without herds of reindeer, he must never- 
theless keep a few head of these animals for winter transport 
where horses are ill adapted to conditions of travel. In the sum- 
mer some nomad friend will allow them to pasture with his own 
herd. Thus, there is a certain dependence of the "agriculturists" 
on the pastoralists; and, even so, existence is possible only because 
the farmer catches fish in the summer.^ 

Similarly, we cannot lightly assume that the economically pro- 
ductive sex or class in a community must be the dominant one. 
Much has been written about the ascendancy of women because 
of their putative invention of farming, but little support is to be 
found for this view. In Uganda a woman with a good plantain 
grove feeds three or four men, but there is no question of femi- 
nine equality.^" On the other hand, the Hopi Indian men grow 
maize, but the women, though in no sense matriarchs, do own the 

* Carl O. Sauer, "American Agricultural Origins: A Consideration of Nature and 
Culture," Essays in Anthropology Presented to A. L. Kroeber (Berkeley, 1936), 
pp. 279-298; A. L. Kroeber, Cultural and Natural Areas of Native North America, 
Univ. Calif. Publ. Amer. Arch, and Ethn., XXXVIII (1939), 146-150. 

*P. L. Smith, Kautokeino og Kautokeino-Lappene (Oslo, 1938), pp. 300-318, 

"John Roscoe, The Baganda (London, 1911), p. 431. 


Economic Factors and Culture 

fields and houses. Evidently there is no simple relation between 
status and productiveness. 

A much more fruitful attack is to envisage the total economic -— ^ 
situation, isolate one factor, and note its correlates. The Ona of 
Tierra del Fuego are guanaco hunters, whose wild vegetable fare 
is negligible, so that women contribute nothing essential to the 
larder. Under the hard conditions of Fuegian life, then, polygyny 
can be only most exceptional, — in contrast to its efflorescence 
where women largely add to the food supply, as in Africa. Even in 
warfare an Ona will generally liberate a captive woman since she 
merely represents an extra mouth to feed.^^ 

Among the Chukchi of northeastern Siberia it is also possible 
to make a worth-while analysis. These people were originally all 
maritime hunters of cetaceans, but in recent centuries a large part 
of the population turned to reindeer-breeding. What, we may i 

ask, has been the effect of this new economic element on the posi- 
tion of woman? The results have been definite, but are not easily ' 
pigeonholed in such rubrics as "superior" and "inferior" status, ' 
for in both groups man is dominant. However, there is a signifi- j 
cant difference: the Maritime Chukchi naturally restrict the peril- 
ous chase of sea-mammals to men; among the pastoralists women 
help tend the herds, hence work harder, but with the advantage ^ 
of being sometimes in the position of owning a herd. Further, 1 
celibacy is rarer, polygyny more frequent, among the nomads. 1 

For while a Maritime Chukchi may find it difficult to support a * ' 

single wife, a Reindeer Chukchi requires a wife for each of his j 

herds. ^^ ' 

Equally suggestive are the consequences of equestrianism * 

among post-Columbian Redskins. The earlier Shoshoneans of 
southern Idaho, like their western congeners, lived in an arid 
country poor in big game, so that salmon and wild roots were their 
staple diet. But the horse introduced by whites revolutionized 
aboriginal economy. It facilitated the killing of large beasts, as 
well as travel across the mountains in search of bison. Because 
of the resulting profusion of hides the skin-dressing industry 

" Martin Gusinde, Die Selk'nam (Modling bei Wien, 1931), passim. 
"Waldemar Bogoras, The Chukchee, AMNH, Memoirs, XI (Leiden, 1909), 


Relation of Ethnology to Other Disciplines 

prospered, and rawhide or other skin containers partly replaced 
basketry. Hitherto these tribes had been riveted to the vicinity 
of stored supplies; now they acquired greater mobility. Thus, 
too, large aggregations of people became possible: it was no 
longer essential to split up into minute family groups for most 
of the year in order to gain subsistence; permanent bands evolved 
as larger politico-social units. The innovation was also potent 
indirectly. Horses lured thieving raiders from the Plains and 
precipitated warfare. This not only fortified band coherence, but 
led to the adoption of sundry Plains features, — skin tents, shields, 
war honors, even new ceremonies. ^^ Thus, the introduction of a 
single transport animal had far-reaching ramifications that could 
not be easily foreseen. 

It is clear, then, that economic factors may be real and potent. 
But, apart from the truism that man must eat in order to live, 
they do not explain the whole or even a preponderant part of 
culture. The point for the scientific investigator is to determine 
precisely how far their influence extends; and that is best done 
by examining the largest possible number of instances in which 
an historically authenticated economic element has been in- 
jected into a culture. 

" Steward, op. cit., pp. 46, 201 ff.; idem, "Changes in Shoshonean Indian Culture," 
Scientific Monthly, XLIX (1939), 524-537. 



Property Rights and Coercive Powers 
of Plains Indian Military Societies 

Among the founders of comparative sociology Hein- 
rich Schurtz deserves a lasting place of honor. Prior to his 
Altersklassen und Mdnnerbiinde (Berlin, 1902) the wide dis- 
tribution and social significance of associations in primitive life 
had escaped the notice of theorists, who confined themselves 
largely to problems connected with marriage, the family, and the 
clan. Schurtz's pioneer effort was far from flawless. There are 
facile generalizations on the psychology of sex and irrelevant 
personal opinions that may be charged to the exhibitionism of 
youth. The author also misinterprets ethnographic evidence from 
sheer ignorance of the facts, which at that time were sadly want- 
ing for many areas. Yet in historical perspective the book com- 
mands respect as one of the landmarks in the development of 
the science.^ 

Journal of Legal and Political Sociology, I (April 5, 1943), 59-71. In somewhat 
different form the second part of this paper was read at one of the Fiftieth Anni- 
versary symposia of the University of Chicago, September, 1941, but has hitherto 
remained unprinted. 

^ For a critique, see R. H. Lowie, Primitive Society (New York, 1920), chap. xi. 


Relation of Ethnology to Other Disciplines 

One of Schurtz's cardinal postulates was the function of age 
diflPerences as creators of social units: age-grades, he contended, 
were humanity's earliest deliberate attempt at social segmenta- 
tion. Thus he naturally came to impress into service the re- 
markable information Prince Maximilian of Wied-Neuwied had 
collected among the Mandan and their neighbors of the upper 
Missouri in the early thirties of the last century.^ For the Prince 
reported from each tribe a series of men's societies differentiated 
by emblems, songs, dances, behavior, and age. Growing older was 
not the sole basis of a higher membership, to be sure, since pay- 
ments were prerequisite. But any thinker who assumed a normal 
sequence of stages in social development could easily dispose 
of the complication: Schurtz simply explained the Plains Indian 
phenomena as transitional from the earlier type of pure age- 
classes to the as yet unattained type of the pure club in wliich 
membership rests solely on payment of a fee.^ 

Field research subsequent to Schurtz and largely stimulated 
by his theories brought out a fact he could not have gleaned 
from Maximilian. Though specific societies were often closely 
paralleled in alien tribes, the pattern found by the Prince as un- 
derlying the entire tribal system of military organizations turned 
out to hold for only a bare half dozen Plains groups. Everywhere 
else societies were neither graded by age nor entered by pay- 
ments. This in itself was no refutation of Schurtz's chronology: 
conceivably the rarer pattern might nevertheless be the older. 
However, the reverse assumption became a priori equally legit- 

More vital was a discovery in part already to be gleaned from 
the earlier sources. A particular organization, such as the wide- 
spread Kit-fox society, was not indissolubly linked with a par- 
ticular age: a young men's company on the upper Missouri, it 
ranked much higher in the Blackf oot ( Montana, Alberta ) scheme, 
and was of course coordinate with other organizations in the 
ungraded systems. Even within a single tribe the place of a 

^ For a summary of these and later data see Clark Wissler, R. H. Lowie, et al.. 
Societies of the Plains Indians, AMNH, Anthropological Papers, Vol. XI. 

^ Heinrich Schurtz, Altersklassen und Mdnnerbiinde (Berlin, 1902), pp. 83flF., 
151-165. , 


Rights and Powers of Plains Indian Societies 

military society varied with the lapse of time. Whereas Maxi- 
milian's Blackfoot in 1833 rated the Dogs as inferior to the 
Ravens, all subsequent observers reverse the order. Similarly, 
Curtis' Braves in the same tribe comprised the oldest bachelors, 
but according to Wissler they were predominantly young mar- 
ried men.'* Membership in association A, B, C, and so on, was 
therefore devoid of an intrinsic, immutable age status. Its basic 
meaning must be sought elsewhere. 

The clue came through the disruptive agency of modern con- 
ditions. Because of the ensuing upheaval, companies of men 
who would normally have jointly acquired a higher society failed 
to buy a new membership. It turned out that these men did not 
automatically rise to a higher grade, but remained in the societies 
they had bought. However, the native attitude went much 
further: in 1910 a Hidatsa (North Dakota) about 90 years old 
considered himself simultaneously a member of organizations 
joined at about seven, twenty, twenty-seven, and forty-five, 
respectively. Identical conceptions obtained among the Mandan 
(North Dakota) and Blackfoot.^ The variations among the 
Arapaho (Wyoming) and Gros Ventre (Montana, Alberta) are 
negligible in this context, for they harmonize with the prin- 
ciple that membership in a graded society of the Plains was 
a form of negotiable property the claim upon which never lapsed 
until it was sold. The very tribes, then, which apparently stressed 
age-stratification in their graded series in actuality considered 
purchase as the basic principle of aflfiliation. 

Whence, however, the empirically established fact that at any 
one time fellow-members were virtually all contemporaries? 
Here Schurtz stands vindicated. He erred, indeed, in asserting 
the necessary priority of associations uniting the age-mates of 
a community, but he was right in ascribing to coevals tlie 
solidarity that could create social units and thereby differentiate 
them from one another. There is abundant evidence from our 
area that young boys were wont to foregather as a spontaneous, 

* Clark Wissler, Societies and Dance Associations of the Blackfoot Indians, AMNH, 
Anthropological Papers, XI (1913), 365-381. Among people who do not reckon 
their ages exactly by years, matrimonial status is significant in this context, as 
Schurtz recognized. 

^ Ibid., pp. 427, 972-975. 


Relation of Ethnology to Other Disciplines 

informal age-class in imitation of their elders. In tribes with a 
system of crystallized age-societies nothing would be more natural 
than for these boys collectively to buy the right to the lowest 
organized society, and later to continue buying successive mem- 
berships at suitable intervals. We must distinguish, then, two 
senses of the term "society." On the one hand, we are dealing 
with sets of incorporeal property rights. Any one complex was 
not intrinsically linked with a particular age grade, so that shifts 
in point of age occurred either through internal rearrangements 
or when a society was borrowed (or bought) by an alien group. 
On the other hand, the gang of united youngsters would regularly 
remain together and subsequently buy their way up to the top 
of the scale. These age-mates, then, do form a body independent 
of their joint property rights; for during the interval between 
their sale of a particular membership and their purchase of an- 
other their sense of solidarity would persist unabated even though 
joint activity might dwindle to a minimum. 

This distinction appears most clearly among the Gros Ventre 
(Montana, Alberta), where each gang bore a unique name but 
shared a set of "dance" privileges with several other gangs, i.e., 
with adjoining age-groups which performed like dances with like 
regalia, but always independently of one another. Thus, a man 
for a lifetime labeled as a Holding-to-a-dog's-tail was reckoned 
a Dog only during and after the performance of the Dog cere- 
mony, sharing that appellation with the other gangs owning the 
same prerogatives.^ 

As to the property rights of the age-societies in general, several 
points are worth noting. As just explained, several groups might 
equally share in the "copyright" to a particular complex. To a 
more limited extent other tribes than the Gros Ventre permitted 
the sharing of the same emblems or other integral parts of com- 
plexes by distinct bodies. This fact might also be phrased by 
saying that the same item could enter distinct complexes. Fur- 
thermore, both the acquisition of membership and the rights ob- 
tained were only in part collective. A Hidatsa gang jointly made 
a large soliciting gift to a higher group in order to coax them into 

* Ibid., p. 933. A. L. Kroeber, Ethnology of the Gros Ventre, AMNH, Anthro- 
pological Papers, I (1908), 232. 


Rights and Powers of Plains Indian Societies 

selling their society; but subsequently each buyer might have 
an individual seller as his ceremonial "father," obtaining by the 
transaction distinctive privileges such as go with honorific offices. 
Finally, there are the tribes with ungraded systems and without 
purchase. Here, e.g., among the Crow (Montana), one gang did 
not replace another, but individuals spontaneously joined or were 
invited to add to the numerical strength of a society. Undeniably 
the members here also had property rights: these could be lost, 
as when one of two rival Crow organizations excelled the other 
in martial exploits and thereby won the right to monopolize the 
losers' tunes for one season. However, from the legal point of 
view, it is an important matter whether possessions are, or are 
not, negotiable: Inalienable land is property in a different sense 
from land that can be disposed of at will. 

The curious point in this connection is that the tribes in ques- 
tion are by no means unfamiliar with the concept of selling in- 
corporeal property, such transactions being in fact a constantly 
recurring phenomenon among the Crow (Montana), who used to 
pay a horse even for such minor privileges as painting the cheeks 
with a certain design. Why, then, did they fail to extend so 
deep-rooted a conception to their military societies? Without 
generalizing for the entire area I should suggest the following 
explanation for the Crow. These Indians did sell a variety of 
privileges, but these come exclusively, or at least preponderantly, 
into the category of sacred possessions, i.e., they were directly or 
indirectly traced to supernatural revelations. But their military 
societies were devoid of religious meaning; a few extremely faint 
reminiscences of visionary experiences to explain the origin of cer- 
tain organizations could not arrest their thorough-going secular- 
ization. The complex of badges, songs, and activities of a military 
society was thus valued quite differently from the corresponding 
complex of the sacred Tobacco societies and fell into a distinct 
category in Crow consciousness. 


For a large portion of our area and adjoining regions to the east 
and north early sources attest a minimum of governmental au- 
thority. Denig's description of an Assiniboine council. Tanner's 


Relation of Ethnology to Other Disciplines 

account of a joint enterprise by Ojibwa (Great Lakes), Cree 
(Central and Western Canada), and Assiniboine (Western Can- 
ada, Montana), Tabeau's observations among the Dakota and 
Arikara (North Dakota), Heame's experiences with his Chip- 
ewyan (Athabaska-Mackenzie area) guide Matonabbee are all 
mutually corroboratory on this point. According to Jones's sum- 
mary of aboriginal Ojibwa conditions, e.g., there were councilors 
with narrowly localized jurisdiction to which status every tribes- 
man was admissible. Their powers were vague and limited, and 
the chief chosen by them was even less able to alter existing 
custom at his pleasure. In Denig's day the head chief of the As- 
siniboine was a purely nominal leader who lacked special pre- 
rogatives and could be humiliatingly overruled by the council. 
The chiefs known to Tabeau were unable to quell a riot and 
might have their authority set at nought by a single resolute 
individual: "insubordination and discord" reigned supreme.'^ 

Making due allowance for exaggerations we find concordant 
testimony, perhaps not for the anarchy implied in some of the 
statements if taken literally, but at least for a marked freedom of 
the individual from physical restraint. This impression is strength- 
ened by the widely held definition of ideal public functionaries. 
The Pawnee (Nebraska) chief, far from being a sovereign ruler, 
was above all a peace-maker and guardian of the village, his 
Hidatsa colleague was "a man of general benevolence who offered 
smoke to the old people and feasted the poor." Their counter- 
part among the Plains Cree was not only expected to exercise 
generosity, but to sacrifice his property for the maintenance of 
order, nay, to forgo vengeance if one of his own kinsmen was 
slain. Correspondingly, a Winnebago (Wisconsin) chief con- 
stantly distributed his possessions and interceded between evil- 

^ E. T. Denig, "Indian Tribes of the Upper Missouri," BAE, 46th Ann. Rept. 
(1930), pp. 431 flF., 435-440, 449; Edwin James, ed.. An Indian Captivity (1789- 
1822): John Tanner's Narrative of His Captivity among the Ottawa and Ojibioa In- 
dians, 1830, Sutro Branch, California, State Library, Reprint Series, No. 20, chap. 
xi, p. 151; Annie Heloise Abel, ed., Tabeau's Narrative of Loisel's Expedition to the 
Upper Missouri (Norman, Okla., 1939), pp. 104-106, 126; William Jones, "Cen- 
tral Algonkin," in Annual Archeological Report for 1905 (Toronto, 1906), p. 137; 
Samuel Hearne, A Journey from Prince of Wales' Fort in Hudson's Bay to the 
Northern Ocean (London, 1795), passim. 


Rights and Powers of Plains Indian Societies 

doers and their revengeful victims; he went so far as to mortify 
his own flesh in order to arouse the pity of the aggrieved, thus 
deflecting their anger from the culprits. In these tribes, then, 
the chief was essentially an appeaser working by cajolery. Small 
wonder that his adjutants and what has been called the camp 
police operated under like restrictions. The duty of all Plains 
Cree of superior status was to prevent strife; and the Black 
Mouths of an Hidatsa village removed misunderstandings, con- 
ciliating aggrieved tribesmen by gentle words and compensatory 
gifts. ^ 

In view of these facts it is startling to find the liberties of the 
Plains Indian periodically suspended by something very much 
like martial law enforced by a body vested with supreme power 
for the time being. Probably the earliest report is Hennepin's. 
In 1680 the explorer met a Santee Dakota (Western Woodlands) 
party, who freely shared their recently obtained supply of 
buffalo meat. Suddenly: 

Fifteen or sixteen Savages came into the middle of the Place 
where we were, with their great Clubs in their Hands. The first 
thing they did was to over-set the Cabin of those that had invited 
us. Then they took away all their Victuals, and what Bears-Oil they 
could find. . . . 

We knew not what these Savages were at first. . . . One of 
them . . . told me, that those who had given us Victuals, had 
done basely to go and forestal the others in the Chase; and that 
according to the Laws and Customs of their Country, 'twas lawful 
for them to plunder them, since they had been the cause that the 
Bulls were all rtin away, before the Nation could get together, which 
was a great injury to the PubHck; For when they are all met, they 
make a great Slaughter amongst the Bulls; for they surround them 
so on every side that 'tis impossible for them to escape.^ 

Tabeau, who was particularly familiar with the Arikara and 
Teton Dakota, tells us that the "soldiers" when elected "to watch 

^ George A. Dorsey and J. R. Murie, Notes on Skidi Pawnee Society, Field Museum 
of Natural History, Anthropological Series, XXVII (1940), 113; R. H. Lowie, 
Notes on the Social Organization and Customs of the Mandan, Hidatsa, and Crow 
Indians, AMNH, Anthropological Papers, XXI ( 1917), 18 f.; David G. Mandelbaum, 
The Plains Cree, ibid., XXXVII (1940), 221, 230, 231 £.; Paul Radin, "The V^inne- 
bago Tribe," BAE, 37th Ann. Kept. (1923), pp. 209 f., 227. 

* Louis Hennepin, A New Discovery of a Vast Country in America . . . ( London, 
1698), pp. 187 f. 


Relation of Ethnology to Other Disciplines 

over the carrying out of the laws of the cerne [bufiFalo surround] 
or over that of some pubhc decision," have the right — only 
temporarily, he emphasizes — "to be severe arbitrarily towards 
every delinquent, to kill his dogs, his horses, to break his weapons, 
to tear the lodges into tatters, and to seize . . . upon all that 
which belongs to him." ^^ This corresponds well to the composite 
picture from a variety of sources on different groups. One would 
merely like to add that though a refractory culprit might be 
severely beaten and even killed, a penitent was rewarded with 
a new lodge and more goods than had been destroyed. Tabeau 
admits other public functions, but is particularly impressed with 
the direction of the collective hunt {cest surtout le cos ou les 
soldats sent severes dans V execution de leur charge) and justifies 
the law as absolutely necessary because to transgress would be 
to "detruire ainsi la base de la suhsistance generale (sic).^' 

It was observations of this type that I once combined with 
principles enunciated by Schurtz into an hypothesis for the 
evolution of the State. My problem was to account for the rise 
of territorial sovereignty from a condition in which, as earlier 
theorists had averred, kinship provided the only bond for joint 
political action. Schurtz himself had already argued that a closely 
knit primitive secret organization could inject order into com- 
munal life such as transcended the power of weak chiefs or was 
precluded by the constant bickering of rival clans. ^^ Among the 
best-known Plains peoples the temporary coercive authority was 
generally vested in military organizations, one or more of which 
units were empowered to punish the crime of stampeding buffalo. 
Thus, these Plains associations seemed a potential instrumentality 
for territorial integration, achieving intermittently, i.e., during 
the surround, what a modern State professes to do continuously. 
Subsequently the hostilities among the several associations within 
a tribe made a strong impression upon me and their mere ex- 
istence appeared as potentially no less disruptive than that of 
contending blood-groups. Nevertheless, I retained the idea that 

'" Abel, op. cit., pp. 116 ff., 245. 

" Heinrich Schurtz, op. cit., p. 363 et passim; R. H. Lowie, Primitive Society; 
idem. The Origin of the State (New York, 1927). 


Rights and Powers of Plains Indian Societies 

the hunt-directing poHce force illustrates "the meteoric display 
of sovereign authority" in "an almost anarchic community." ^^ 

Naturally the question obtrudes itself why an approach to 
complete freedom should thus alternate with subjection to co- 
ercive force. For this there readily suggested itself an answer 
in economic terms. Securing an ample supply of buffalo meat 
was a matter of life and death, hence under the threat of starva- 
tion the Indians willingly surrendered their normal rights, sub- 
jecting themselves to a rigorous discipline. 

In recent years Drs. Hoebel and Provinse have independently 
criticized my position.^^ They have not, however, rejected my 
specimens of compulsion as spurious, nor do they consider them 
irrelevant to the problem of political development. The gist of 
their comments is rather that I have understated the case: 
"Sovereignty" — to use a grandiloquent term which I invest with 
no fetichistic reverence in this context — is, they argue, less 
sporadic and more inclusive than my exposition suggests. It was 
misleading to overemphasize the spectacular disciplinary con- 
comitants of the buflFalo hunt. My own field data, as well as those 
of other observers, are aptly cited to show that the same pro- 
cedures held in several distinct circumstances. 

I accept the criticism as valid and should like to strengthen 
it by additional evidence. At the same time I must qualify some 
of its implications. 

Ethnographic facts that contravene the unique significance 
of the buffalo surround come from the area to the east of the 
Plains. As Skinner and Macleod have noted,^^ several Woodland 
peoples, i.e., peoples for whose economy buffalo-hunting was a 

" Lowie, The Origin of the State, pp. 107 fF., 116. 

" E. Adamson Hoebel, "Associations and the State in the Plains," American 
Anthropologist, XXXVII (1936), 433-438; idem, The Political Organization and 
Law-Ways of the Comanche Indians, AAA, Memoirs, No. 54 ( 1940), p. 82; John H. 
Provinse, "The Underlying Sanctions of Plains Indian Culture," in Fred Eggan, 
ed., Social Anthropology of North American Tribes (Chicago, 1937), p. 365. 

" Alanson Skinner, Social Life and Ceremonial Bundles of the Menomini Indians, 
AMNH, Anthropological Papers, XIII (1913), 22-26; idem. Material Culture of 
the Menomini (New York, 1921), pp. 51 f.; idem. Political Organization, Cults, and 
Ceremonies of the Plains — Ojibway and Plains Cree Indians, AMNH, Anthropologi- 
cal Papers, XI ( 1914), 498 f.; Radin, op. cit., pp. 114, 209, 220, 226 f. 


Relation of Ethnology to Other Disciplines 

subordinate or even negligible feature, paralleled the phenomena 
so strikingly manifested by their western neighbors. The Win- 
nebago constabulary did control the chase, but — like the Me- 
nomini (Wisconsin) counterpart — they also proceeded in similar 
fashion to forestall premature exploitation of wild rice. The Sauk 
and Fox (Wisconsin) "war chiefs" directed not the hunt itself, 
but the homeward journey from it, their aim being to preclude 
hostile attacks on single families and the pillaging of corn by 
nimble marauders. Every night one war chief would set up liis 
staff as a boundary mark, and whoever stepped beyond it had 
"his canoe and whatever else he may have along with him 
destroyed." As soon as the village routine was restored this 
martial law lapsed; in no other circumstance did our authority 
discover "any laws enforced or penalties exacted for disobedience 
of them." The Winnebago police had a variety of functions be- 
yond those already mentioned. Besides preventing a stampede 
of game and premature inroads on wild rice, they regulated travel 
in the Sauk and Fox fasliion, guarded the village continuously, 
and whipped seducers of women. Oddly enough, in cases of 
murder these coercers figured as appeasers. 

It seems idle to speculate whether the police institution as we 
find it in the Plains originated there or in the Woodlands, 
especially since we know that many of the "Plains" tribes emi- 
grated to their historic homes from the forested regions in 
relatively recent times. But it is evident that coercive functions 
are not indissolubly tied up with the conditions of hunting herds 
of big game. As indicated they coexist with radically distinct 
pursuits. Further, they do not automatically arise from the 
buffalo surround. For the Sarsi (Alberta), the Comanche (Texas), 
the Shoshone (Idaho, Wyoming), marginal but in many ways 
thoroughly acclimatized representatives of the Plains, hence 
devoted buffalo hunters, either lacked punitive measures or 
reduced them to a minimum. ^^ 

Hoebel's and Provinse's criticisms are therefore well taken: tlie 

^^ Diamond Jenness, The Sarcee Indians of Alberta, National Museum of Canada, 
Bull. 90, Anthropological Series, No. 23 (1938), pp. 11, 41; Dimitri Shimkin, per- 
sonal communication concerning the Wind River Shoshone; Hoebel, The Political 
Organization . . . of the Comanche Indians, p. 82. 


Rights and Powers of Plains Indian Societies 

range of coercive powers did not coincide with buffalo hunt- 
ing but embraced various other aspects of tribal life in the 
Plains and adjoining regions. As already indicated, this revision 
involves at bottom not a refutation but a strengthening of my 
basic contention, for it asserts a far broader base than the narrow 
economic one envisaged by me. 

Though this is very gratifying, I must turn devil's advocate 
against the possible implication that the police phenomena dem- 
onstrate more than incipient Statehood. On the one hand, as 
explained above, the genuinely authoritarian aspects of various 
tribal constabularies (Crow, Winnebago) are superseded by 
purely persuasive functions in so vital a crisis as intratribal murder. 

Secondly, we must recognize the seasonal dichotomy of social 
life among many Plains tribes, roughly paralleling the contrasts 
emphasized by Durkheim for the Australians and by Mauss for 
the Eskimo. ^^ 

Wissler stated the facts long ago,^^ pointing out the intermittent 
character of governmental control in our area, but my previous 
publications fail to give due emphasis to the data. A few rep- 
resentative phenomena may be cited. The Blackfoot held their 
council in the summer, separated after the fall hunt, and re- 
assembled in the spring. Hidatsa villages were under a winter 
chief whose term began in autumn and ended with the melting 
snow. The Dakota and the Crow reorganized their military 
societies every spring and these companies functioned until 
the first snowfall. Three recent monographs bring out the essential 
point with startling clarity. For the Cree, the Sarsi, and the 
Kiowa (Oklahoma) social life culminated in the annual Sun 
Dance. It was for that ceremony that Cree bands would unite 
in the summer, soon thereafter dispersing and finally, in mid- 
winter, breaking up into minute family groups too small to 
permit associational activity. Similarly, the Sarsi hunted buffalo 
in small tribal segments, and only for the Sun Dance one society, 
the Red Paint organization, assumed control. Again the Kiowa 

" Marcel Mauss, "Essai sur les variations saisonnieres des soci^tes Eskimo: fitude 
de morphologie sociale," L'Annee sociologique, IX (1906), 39-132; Emile Durk- 
heim, Les Formes elementaires de la vie religieuse (Paris, 1912), passim. 

^' Clark Wissler, The American Indian (New York, 1922), pp. 161, 178. 


Relation of Ethnology to Other Disciplines 

associations "functioned only during the four or so weeks of the 
Sundance gathering." ^^ 

The lapse of authoritarianism for a considerable, or in some 
tribes even the major, part of the year as a result of economic 
conditions and consequent modes of settlement, inevitably 
reduces whatever sovereignty exists to a nascent stage. 

Finally, as Provinse aptly shows, there may be dispersal of 
authority between two or even more agencies. For example, 
normally during the season of reunion the Crow camp would be 
under the guidance of a chief and the military society appointed 
by him; but at a Sun Dance the director of that ceremony is 
supreme and appoints another association to serve for its dura- 
tion. ^^ Such temporal alternation of supreme power evidently 
militates against the centralization of authority. 

However suggestive, then, the data on police functions in and 
near our area are, they suggest nascent rather than achieved 
governmental integration. The failure to extend the coercive 
authority of police associations over the entire field of internal 
relations (notably in the case of murder); the merely periodic, 
in some cases even ephemeral, assumption of such authority; 
its dispersal between two or more foci; the disruptive tendency 
of strife between rival associations pointed out in The Origin 
of the State — all militate against the creation of a full-fledged 

Nevertheless, the potential jurisdiction of a military society 
remains a most significant fact for our theme. This has been very 
effectively demonstrated for the Cheyenne (Montana). Al- 
though here, too, the seasonal dichotomy breaks up the societies 
during the fall and winter, this did not apply to the Dogs, who 
by an historical accident coincided with one of the tribal bands, 
hence preserved a unique solidarity. In fact, some semblance of 
cohesion obtained even in the other societies inasmuch as 
members who happened to reside together during the period 

" Idem, The Social Life of the Blackfoot Indians, AMNH, Anthropological Pa- 
pers, VII ( 1911 ), 22-26; idem. Societies and Ceremonial Associations in the Oglala 
Division of the Teton-Dakota, AMNH, Anthropological Papers, XI (1912), 24; 
Lowie, Notes on the . . . Mandan, Hidatsa, and Crow Indians, pp. 18 f.; Mandel- 
baum, op. cit., pp. 203, 225; Jenness, op. cit., pp. 11, 41 f.; Jane Richardson, Law 
and Status among the Kiowa Indians ( New York, 1940 ) , pp. 9 f . 

" R. H. Lowie, The Crow Indians (New York, 1935), p. 308. 


Rights and Powers of Plains Indian Societies 

of dispersal might join in action as occasions arose. Possibly 
most illuminating is a case of mayhem. Among Plains Indian 
peoples generally such an offense was only a tort, the police as- 
suming the function of mere go-betweens in trying to persuade 
the aggrieved party to accept an idemnity. Here, however, some- 
thing radically different occurred: the Foxes then serving as 
police treated the matter as a public wrong, severely beat the 
culprit, and accepted the "fine" offered by him without sharing 
it with his victim.^*^ This certainly suggests that the governmental 
powers of a military organization were capable of very consider- 
able enlargement. 

One question which remains is whether the germs of sov- 
ereignty were peculiarly tied up with the rise of military as- 
sociations. Reverting to the ethnographic correction made above 
concerning the supposedly unique effect of the buffalo surround, 
I should like to point out that the geographical extension of our 
police phenomenon automatically supplies an answer. For in the 
Woodlands there are no military societies of either the graded or 
ungraded type. Moreover, it has long been recognized that 
the conventional "Plains Indians" comprise a congeries of cul- 
turally disparate groups. Some are matrilineal, others patrilineal, 
still others without definite rules of descent. In one sector of 
the area kinship nomenclature is of the Omaha type, almost 
exactly duplicated in the western Woodlands; in another we find 
its precise logical antithesis, whose closest parallels crop up in 
the Southeast and in Arizona; still another province displays a 
third pattern. Agriculture is totally unknown to some tribes, 
rudimentary among others, more highly developed among the 
Pawnee. The Sun Dance, the dominant festival of the high 
Plains, dwindles to insignificance or disappearance among South- 
em Siouans, who in part substitute an equivalent of the Wood- 
land ceremonial dramatizing the ritual killing and revival of 
initiates. These facts long ago recognized in conversation by the 
late Alanson Skinner have at last been adumbrated taxonomically 
by the suggestion of a "Wisconsin-Prairie" area.-^ 

^ K. N. Llewellyn and E. A. Hoebel, The Cheyenne Way (Norman, Okla., 1941 ), 
chap. V, esp. pp. 99-101, 110 f., 115, 118 f., 122 ff. 

" A. L. Kroeber, Cultural and Natural Areas of Native North America, Univ. 
Calif. Publ. Amer. Arch, and Ethn., XXXVIIl (1939), 85. 


Relation of Ethnology to Other Disciplines 

Now within the Prairie subdivision of this area we discover a 
feature shared not only with the complementary Wisconsin 
sector, but with a variety of still more easterly Woodland tribes: 
public functions of one sort or another devolve on the clan. 
Thus, it is the Bear clan that policed a Winnebago village, where, 
incidentally, the chief as an appeaser belonged to another, the 
Thunderbird, clan. This does not necessarily mean that all adult 
men of a certain clan formed the permanent police or that all 
the members of a squad were clansmen; the point is rather 
that at least the responsibility for recruiting and supervising the 
police rested upon one or more specific clans. Thus, the two 
Osage (Missouri) chiefs, each representing one of the two 
moieties and a particular "gens," i.e., patrilineal clan, within it, 
appointed the marshals for the hunt, one from each of certain 
gentes; and honorific titles devolved on three of the officers, 
each representing one of these units. Similarly, among the Iowa 
"the Elk gens furnished the soldiers or policemen"; and according 
to Fletcher and La Flesche, two clans were associated with the 
regulation of the hunt among the Ponca ^^ ( Nebraska ) . 

The recruiting of a constabulary from definite clans is not, 
however, the only alternative to identifying it with a particular 
military association. Several Plains tribes used the device of what 
may be called a "nonce police." The Pawnee, for example, 
though vesting disciplinary powers within the village in the 
hands of the chiefs adjutant and three appointees of his, reg- 
ulated the bufi^alo hunt on a different principle, a priest select- 
ing one of four possible organizations for that particular enter- 
prise only. So the Omaha council would delegate to men from 
the class of brave warriors the task of controlling some com- 
munal hunt. The appointees formed no permanent body, but 
owing to their fitness were likely to be subsequently drafted 
for police duties in the village. From sources on the Plains Ojibwa, 
Dakota, and Assiniboine it seems probable that wherever men 
of recognized valor formed a distinct class they were the obvious 

^ Francis La Flesche, "The Osage Tribe," BAE, 36th Ann. Kept. ( 1921 ), pp. 66- 
68; J. O. Dorsey, "Siouan Sociology," BAE, 15th Ann. Rept. (1897), p. 238; AUce 
C. Fletcher and Francis La Flesche, The Omaha Tribe, BAE, 27th Ann. Rept. 
(1911),pp. 45, 210, 279. 


Rights and Powers of Plains Indian Societies 

candidates for constabulary duty. In short, any preexisting unit 
clearly defined in native consciousness, whether military club, 
clan, or distinguished warrior group, readily lent itself to such 

Summarizing the above remarks, I still feel that the military 
organizations of the Plains area exemplify the potentialities of 
associations as regards the creation of supreme central authority. 
It is merely necessary to remember that such germs of Statehood 
actually as a rule remained rudimentary, though the Cheyenne 
case demonstrates that some tribes carried them much nearer 
to fruition than others. Further this type of unit is obviously not 
peculiarly fit to absorb disciplinary functions. Where military 
societies are lacking, such activities quite as naturally devolve 
on other preexisting units, such as clans or a general honorary 
class of braves. 

Note. — For the orientation of the reader I hst alphabetically the location 
of Indian tribes mentioned in the article. Where the habitat changed re- 
peatedly in historic times, one or two significant locations are taken to 
suflBce for present purposes. 

Arapaho — Colorado, Wyoming Hidatsa — N. Dakota 

Arikara — N. Dakota Iowa — Iowa 

Assiniboine — Montana, Alberta Kiowa — Oklahoma 

Blackfoot — Montana, Alberta Mandan — N. Dakota 

Cheyenne — Montana Menomini — Wisconsin 

Chipewyan — Hudson Bay to Ojibwa — Great Lakes Region 

Lake Athabaska Omaha — Nebraska 

Comanche — Colorado to Texas Osage — Missouri 

Cree (Plains) — Manitoba to Pawnee — Nebraska 

Alberta Ponca — Nebraska 

Crow — Montana Sarsi — Alberta 

Dakota (Santee) — Minnesota Sauk — Wisconsin 

Dakota (Teton) — S. Dakota Shoshoni — Wyoming, Idaho 

Fox — Wisconsin Winnebago — Wisconsin 
Gros Ventre — Montana, Alberta 

"' Dorsey and Murie, op. cit., p. 113. J. O. Dorsey, "Omaha Sociology," BAE, 
3d Ann. Kept. (1884), pp. 233 £., 288, 321, 363; Fletcher and La Flesclie, op. cit., 
pp. 210, 279; James, op. cit., chap. xii. Denig, op. cit., p. 436; R. H. Lowie, So- 
cieties of the Crow, Hidatsa and Mandan Indians, AMNH, Anthropological Papers, 
XI (1913), 132-136. 



Some Aspects of Political Organization 
Among the American Aborigines 

Huxley Memorial Lecture For 1948 

In a gross description of continental areas the Ameri- 
can aborigines figure as separatistic and democratic, contrasting 
in the former respect with the African Negro, in the latter with 
both African and Polynesian. The illuminating studies on African 
politics edited by Drs. Fortes and Evans-Pritchard have demon- 
strated decisively what readers of P. A. Talbot or Henri Labouret 
had long known, to wit, that the traditional picture of Negro 
government is over-simplified. To be sure, there have been many 
powerful monarchies in African history, but east of the Niger, 
in the Upper Volta region, and in the Anglo-Egyptian Sudan 
not a few tribes resist integration as much as any people in 
the world. In 1931 the 69,484 Lobi on French soil in the Upper 
Volta country were spread over 1,252 mutually independent sham 
villages (pretendus villages); a single one had over 600 residents, 
while 44 of these hamlets numbered fewer than 100, so that M. 
Labouret properly speaks of a particularisme accuse. Within no 

Journal of the Royal Anthropological Institute of Great Britain and Ireland, 
LXXVIII (1948), 11-22. 


Political Organization of American Aborigines 

unit were there any chiefs, and assembhes convened to adjudicate 
particular issues had no means to execute their decisions. In 
short, the gamut of possible variations is realized in Negro 
Africa: we find there vast kingdoms on the pattern of Uganda 
and Benin, but also minute, headless, "anarchic" groups.^ 

In the present essay I shall examine the corresponding phe- 
nomena in aboriginal America. In a discussion of this sort it is 
convenient, if not inevitable, to use such terms as "the State," 
"law," "government," "political," "sovereignty." Conforming to 
the views of Max Weber, Professor Radcliffe-Brown, and Profes- 
sor Thurnwald as I understand them, I take these words to imply 
the control of physical force so far as a given society recognizes 
it as legitimate. Thus, the King of Uganda could rightfully order 
the execution of a subject, no matter how arbitrary the decree 
might seem from our point of view; and in West Africa the 
Mumbo Jumbo organization properly flogged malefactors. On 
the other hand, similar acts by the Ku Klux Klan are in usurpa- 
tion of functions monopolized by the State in Western civili- 

However, a genetic view of political structure must reckon 
with the fact that primeval anarchy could not suddenly blos- 
som forth into a modern State claiming absolute dominance 
within its territorial limits. It is, indeed, a documented fact that 
the states of the most advanced modern peoples did not develop 
contemporary pretensions until relatively recent times, yet their 
immediate antecedents did have a political organisation, in other 
words, laws and government. A simple society may be differen- 
tiated so as to foreshadow government, yet the coercive element 
may be lacking. The Yurok of north-western California and the 
Ifugao of Luzon have no chiefs or judges whatsoever, yet a 
dispute in their midst is settled by unoflBcial go-betweens ap- 
proved by public opinion, who offer their services, though with- 
out an iota of authority. A logical dichotomy of societies on the 
rigid definition of Statehood indicated above would rule out 

^ Meyer Fortes and E. E. Evans-Pritchard, eds., African Political Systems (Ox- 
ford University Press, 1940); Henri Labouret, Les Tribus du Rameau Lobi, Uni- 
versite de Paris, Travaux et Memoires de I'lnstitut d'Ethnologie, XV (1931), 56, 
215, 386. 


Relation of Ethnology to Other Disciplines 

such phenomena as quite irrelevant to a study of government, 
but the common sense of comparative jurists regards them as 
highly significant. In the following inquiry, then, I shall indeed 
retain the exercise of force as the criterion of a full-fledged 
political organization, but I shall also consider what seem evolu- 
tionary stages toward that consummation. 

The questions I ask concerning American Indians may be 
phrased as follows: Within what territorial limits does authority 
create some measure of solidarity? And what is the nature of 
the authority encountered? Specifically, where, in America, was 
a state of modern type realized? What trends can be discerned 
toward its evolution? 


Notwithstanding my initial qualifications, African systems on 
the whole do differ noticeably from those of the New World. 
According to Roscoe, the Baganda once numbered three million; 
by 1911 civil wars and the sleeping sickness had sadly reduced 
them, but not below the million mark. In 1668 Dapper credited 
Benin with a regular army of 20,000, which at a pinch could be 
increased to five times as many; the capital was five or six 
Dutch miles in circumference and had thirty main streets. In 
about 1870 Schweinfurth set the Shilluk at over a million; partly 
because of wars recent estimates are far more modest, yet they 
fluctuate between 50,000 and 100,000. Shortly before this ex- 
plorer's visit a million Mangbettu had been under the sway of 
a single ruler. More recently the king of Ashanti had a quarter of 
a million subjects.^ 

Except in the few higher civilizations of Mexico, Yucatan, 
Colombia, and Peru, there is nothing to match even the least of 
these figures, apparent parallels proving deceptive. To be sure, 
aboriginal Chile is said to have been inhabited by from half 
a million to a million and a half Araucanians, but "there was no 
peacetime overall chief, no centralization of authority." There 
were, indeed, greater and lesser territorial units, but the sub- 
ordination of the smaller "must have been close to purely nomi- 

=■ John Roscoe, The Baganda (London, 1911), p. 6; P. A. Talbot, The Peoples of 
Southern Nigeria (Oxford University Press), pp. 162 f.; George Schweinfurth, 
The Heart of Africa ( London, n.d. ), I, 15, II, 35. 


Political Organization of American Aborigines 

nal." Only during the nineteenth century "the earher atomistic 
peacetime poHtical structure assumed somewhat greater unity, 
cohesion and hierarchization." To take a humbler figure, the 
55,000 Navaho now rank as the largest native tribe in the United 
States. But, in the first place, theirs has been a mushroom growth: 
in 1868 they did not exceed 15,000 — ^possibly not 9,000. Secondly, 
it is not clear that even this number were ever under a single 

As a matter of fact, a tendency to separatism was general. So 
advanced a people as the Hopi — some 3,000 in all — live in eleven 
villages, mislabelled "towns" by grandiloquent ethnographers. 
Yet even this paltry population neither has nor has had a com- 
mon head: "between pueblo and pueblo there is an attitude of 
jealousy, suspicion and subdued hostihty." '^ 

Much ado has been made about the Creek Confederacy in 
the south-eastern United States and the Iroquois League of 
northern New York State. Unquestionably both prove wider 
political cooperation than was common in the New World, but 
their achievements must not be overrated. Authenticated oc- 
currences reduce the cohesion involved in these alliances to a 
proper scale. It so happened that one of the Creek tribes, the 
Kasihta, became friendly with the alien Chickasaw. When the 
latter were at war with the Confederacy in 1793, "the Kasihta 
refused to take up arms with the other Creeks and their right 
to act in this independent manner was never questioned." Strictly 
parallel conduct among the federated Iroquois during the Ameri- 
can Revolution was noted by Morgan. Each tribe was permitted 
to decide upon its course of action: the Oneida and half of the 
Tuscarora sided with the colonists, the other "leagued" tribes 
with the English. It was as though in 1914 Bavaria and half of 
Baden had joined the Allies to fight their fellow-Germans. Apart 
from this disintegration in a crisis, earlier claims on behalf of 
the League's influence have been exploded by Fenton's his- 
torical researches. The Iroquois did raid far and wide, but it 
hardly holds true that "their dominion was acknowledged from 

^ John M. Cooper, "The Araucanians," Handbook of South American Indians, 
BAE, Bull. 143, Vol. II ( 1946), 694, 724; Clyde Kluckhohn and Dorothea Leighton, 
The Navaho (Cambridge, 1946), pp. xv, 73. 

* Mischa Titiev, Old Oraihi, PMH, Papers, Vol. XXII, No. 1 ( 1944), 59-68. 


Relation of Ethnology to Other Disciplines 

Ottawa river to the Tennessee and from the Kennebec to Ilhnois 
River and Lake Michigan." In any case, at its peak in the 
seventeenth century the League never embraced over 16,000 or 
at most 20,000 persons.^ 

Since the one-eyed is king among the bhnd, the two falter- 
ing attempts at consohdation by the Creek and the Iroquois re- 
main noteworthy "cHmactic" results, as my colleague Professor 
Kroeber might phrase it. In world perspective, however, they 
are unimpressive. 

If skilful farming populations showed no greater sense of 
nationalism, little can be expected of the hunters. The Caribou 
Eskimo lacked permanent political units altogether, each com- 
munity being in Professor Birket-Smith's judgment "an incoherent 
conglomerate of families or households, voluntarily connected 
by a number of generally recognized laws." The largest settle- 
ments have a population of about 50, and all of them jointly do 
not exceed ten times that figure. Earlier reports, to be sure, sug- 
gest a recent decline, due largely to famine, but even half a 
century ago the largest separate tribe of the area was not credited 
with over 178 souls. To turn toward the southern tip of the New 
World, the Ona population at its peak is set at between 3,500 
and 4,000. Since this embraced 39 wholly independent territorial 
hordes, the average size of the political unit was about 100.^ 

Extreme as the Eskimo and the Fuegian instances may seem, 
they are paralleled on varying levels of cultural complexity. The 
exceptionally favourable food supply of North-west Californians 
failed to produce solidarity beyond the bounds of kinship and 
of immediate proximity. Of the seventeen independent Yurok 
hamlets listed in 1852, the largest had only 165 inhabitants; three 
others had over 100; five, well under 50. 

Up and down the Pacific coast of North America similar con- 

^ J. R. Swanton, "An Indian Social Experiment and Some of Its Lessons," Scientific 
Monthly, XXXI ( 1930), 368-376; J. N. B. Hewitt, "Iroquois," Handbook of Amer- 
ican Indians North of Mexico, BAE, Bull. 30 (1907); V^illiam N. Fenton, "Prob- 
lems Arising from the Historic Northeastern Position of the Iroquois," Smithsonian 
Miscellaneous Collections, Vol. C (1940); L. H. Morgan, Ancient Society (New 
York, 1877), Part II, chap. 5. 

" Kaj Birket-Smith, "The Caribou Eskimos," Report of the Fifth Thule Expedition, 
1921-1924 (Copenhagen, 1929), Vol. V, Part I, pp. 65-75, 260; Martin Gusinde, 
Urmenschen im Feuerland (Berlin, Wien, Leipzig, 1946), p. 97. 


Folitical Organization of American Aborigines 

ditions prevailed. In north-eastern Washington something less 
than 1,500 Sanpoil were spread over twenty villages, each of 
which, except for those conspicuously small, was autonomous. 
The Quinault, in the south-western part of the same state, prob- 
ably numbered 800, divided among roughly 20 villages. The 
Lemhi of Idaho and associated Shoshoneans are set at 1,200 about 
the beginning of the nineteenth century, and this included more 
than a single group; Lewis and Clark estimated one group at 
100 warriors and 300 women and children; another at 60 warriors. 
In eastern Brazil the Botocudo stock was split into several distinct 
tribes, some of them subdivided into bands from 50 to 200. 
Notwithstanding the existence of tribal chiefs, an authority re- 
ports "the constancy of their blood feuds, not only between dis- 
tinct tribes, but even between bands of the same tribe." The 
Foot Indians of the Gran Chaco gathered in bands approximat- 
ing the Botocudo pattern."^ 

No doubt an intermediate order of magnitude occurred. The 
Cheyenne of the Northern Plains at one time probably numbered 
not far from 4,000. Of the Ge stock, some members were in- 
considerable enough: the recent Canella fluctuated about the 
300 mark, but earlier travellers describe the villages of their con- 
geners as rather larger. In 1824, for example, one Apinaye settle- 
ment had a population of 1,400; and the more remotely related 
Sherente display a sense of solidarity beyond the immediate local 
group. Though a paramount head is wanting, the several village 
chiefs sometimes jointly depose a grossly deficient colleague and 
appoint his successor. Characteristically, however, the Sherente 
have long been at bitter enmity with the Shavante, their closest 
linguistic and cultural kin.^ 

'A. L. Kroeber, Handbook of the Indians of California, BAE, Bull. 78 (1925), 
p. 16; Verne F. Ray, The Sanpoil and Nespelem: Salishan Peoples of Northeastern 
Washington, University of Washington Publications in Anthropology, V ( 1932 ) , 
21-24, 109; Ronald L. Olson, The Quinault Indians, ibid., VI ( 1936), 22; Julian H. 
Steward, Basin-Plateau Aboriginal Sociopolitical Groups, BAE, Bull. 120 (1938), 
pp. 188 f.; Curt Nimuendaju, "Social Organization and Beliefs of the Botocudo of 
Eastern Brazil," Southwestern Journal of Anthropologij, II (1946), 97 ff.; A. Me- 
traux, "Ethnography of the Chaco," Handbook of South American Indians, BAE, 
Bull. 143, Vol. I ( 1946), 302; idem, "The Botocudo," BAE, Bull. 143 ( 1946), p. 536. 

*K. N. Llewellyn and E. A. Hoebel, The Cheyenne Way (Norman, Okla., 1941), 
p. 78; Curt Nimuendaju, The Apinaye, The Catholic University of America, Antliro- 


Relation of Ethnology to Other Disciplines 

Similar qualifications apply to the instances from the eastern 
United States. The League of the Iroquois has already been dis- 
cussed. The Cherokee and the Choctaw were the two largest 
south-eastern tribes, being estimated at 22,000 and 15,000 souls, 
respectively, in 1650. However, once more the political unit is 
incomparably smaller than the linguistic. For the Choctaw, 
Swanton reasonably suggests some 40 to 50 synchronous com- 
munities "constituting small States, each with its chief." An 
anonymous French writer of ca. 1755 does speak of a grand chef 
of the nation, but adds that his authority was negligible. The 
Cherokee were scattered over at least 80 towns. "These people 
came under the domain of one tribal chief only in times of 
great emergency and then most imperfectly." On the whole, it 
seems likely that the figures set for the Natchez in 1650 and for 
the Powhatan in 1607 — 4,500 and 9,000 — approach the limits 
attained within the area by any governmental entity.^ 

At this point it is well to recall the phenomenon luminously 
illustrated by Durkheim for Australians, by Mauss for the Eskimo, 
and since demonstrated elsewhere. The seasonal rhythm of life, 
rooted in economic exigencies, transforms the constitution of a 
group and, as a corollary, its social life. The consequences we 
shall consider later. For the present, we merely note that some 
of the figures quoted would hold only for a relatively brief por- 
tion of the year; at other times, the tribe breaks up into minute 
fragments in order more effectively to exploit the environment. ^° 

To review the argument, American figures of a population ap- 
proximating or exceeding 10,000 rarely, if ever, refer to per- 
manently integrated political units. 

How far does this conclusion apply to the four higher civiliza- 
tions? As for the Aztec, the moot question of whether they totaled 

pological Series, No. 8 (Washington, 1939), p. 7; idem, The Serente, Publications 
of the Frederick Webb Hodge Anniversary PubUcation Fund ( Los Angeles, 1942 ) , 
IV, 9 f . 

"W. H. Gilbert, The Eastern Cherokees, BAE, Bull. 133 (1943), p. 363; J. R. 
Swanton, Source Material for the Social and Ceremonial Life of the Choctaw 
Indians, BAE, Bull. 103 (1931), pp. 90, 95, 243; idem. The Indians of the South- 
eastern United States, BAE, Bull. 137 ( 1946), pp. 114, 123, 161, 175. 

^'' Emile Durkheim, Les Formes elementaires de la vie religieuse (Paris, 1912); 
Marcel Mauss, "Essai sur les variations saisonnieres des societes Eskimo: fitude de 
morphologic social e," L'Annee sociologique, IX (1906), 39-132. 


Political Organization of American Aborigines 

three or many more millions need not concern us; we are in- 
terested solely in what number belonged to the same state. That 
the hoary idea of an Aztec empire is untenable seems certain 
in the light of modern research. All we find is a belated league of 
three tribes which remained mutually distrustful: "the Aztecs 
had no sense of unity," no national spirit. Within the present limits 
of Mexico City, Tenochtitlan and Tlatelolco long persisted in 
complete independence of each other. At the time of the Spanish 
invasion the Texcocans joined the intruders against their former 
ally, Tenochtitlan. A quarter of a million people, or thereabouts, 
may possibly have had a single government on a strict definition. ^^ 

Maya ruins are spread from northern Yucatan to Honduras, 
but they belong to difi^erent periods, and it is not easy to estimate 
the residents of any one state. Possibly in about a.d. 1000, ac- 
cording to legendary history, there was a league of three cities, 
of which Mayapan gained the ascendancy, establishing a cen- 
tralized government two or tliree centuries later. This was 
followed by disintegration, leaving only petty chieftains for the 
Spaniards to contend with. In their era the rulers of Mani were 
"the most powerful in Yucatan." The tribute list for that province 
demonstrates 13,480 adult males. If we multiply this by six, or 
even ten, we still get no total population that looks spectacular 
by an African scale. -^^ 

The Chibcha numbered possibly a million, but they too were 
divided up among several distinct states, of which Zipa, the 
largest, is credited with 300,000 souls. The untrustworthiness of 
early estimates is indicated by a fantastic reference to armies 
of 50,000 whereas no more than 600 Zipa braves attacked the 
Spanish troops. ^^ 

In short, the solitary convincing instance of grandiose ex- 
pansion in the Western Hemisphere is that of the Incas of Cuzco, 
Peru. Their realm did extend from Ecuador to northern Chile, 

"G. C. Vaaiant, Aztecs of Mexico (Garden City, 1941), pp. 91, 134, 213 f. 

" A. M. Tozzer, Landa's Relacion de las Cosas de Yucatan, PMH, Papers, XVIII, 
64; Sylvanus G. Morley, An Introduction to the Study of Maya Hieroglyphs, BAE, 
Bull. 57 ( 1915), pp. 2-12; Ralph L. Roys, The Book of Chilam Balam of Chumayel, 
Carnegie Institution of Washington, Publ. 438 (1933), pp. 188-195. 

^* A. L. Kroeber, "The Chibcha," Handbook of South American Indians, BAE, 
Bull. 143, Vol. II (1946), 887-909. 


Relation of Ethnology to Other Disciplines 

embracing possibly 6,000,000 subjects. However, we must rec- 
ollect that aggrandizement was a very late pre-Columbian 
achievement. "In early times neither the Inca nor any of their 
neighbours thought of organizing their conquests as a permanent 
domain." Until the reign of Pachacuti (ca. a.d. 1438) "towns 
very near to Cuzco preserved complete freedom of action and 
raided one another's territory whenever there seemed to be a 
good opportunity for plunder." ^* 

With a unique exception, then, the American Indians must 
be regarded as eminently separatistic. 

However, there was certainly no sudden mutation from an 
Ona-like to an Inca-like condition. The Creek and the Iroquois 
schemes indicate a stage of solidarity, however imperfect, on 
a larger than normal scale. Still more illuminating are phenomena 
within the historic period. Whereas the two well-known leagues 
united mainly communities of like or closely related speech, 
Pontiac (1763) and Tecumseh (died 1813) brought together 
wholly unconnected tribes. The Ottawa chief rallied not merely 
his own people and their Algonkian congeners, but also the 
Seneca and the Wyandot of Iroquoian stock and the Siouan Win- 
nebago. The Shawnee leader arrayed Algonkians, Wyandot, and 
even Creek Indians against the United States. Though both up- 
risings proved abortive, though they culminated in negation of 
British and American overlordship rather than in the creation of 
a close-knit aboriginal state, they do prove that under strong 
emotional stimulus exceptional natives could and did visualize 
cooperation of major scope. Individuals of comparable organiz- 
ing skill, however diverse their motivation, must be credited 
with the nascent forms of Andean imperiaUsm.^^ 

I now turn to my second theme — the manifestation or adumbra- 
tion of coercive authority in aboriginal America. As in Africa, so 
here too, the range of observable phenomena is very great. At 
one extreme we find the "anarchic" Eskimo, north-west Califor- 

" John H. Rowe, "Inca Culture at the Time of the Spanish Conquest," ibid., 
pp. 184 f., 201-209, 257 S. 

^^ James Mooney, "The Ghost Dance Religion and the Sioux Outbreak of 1890," 
BAE, 14th Ann. Kept. (1896), pp. 681-691. 


Political Organization of American Aborigines 

nians, and Fuegians; at the other, the Incas of Peru. But in the 
New World, the latter must be regarded as atypical, and an 
intermediate condition represents the norm. By this I mean a 
condition with diflPerentiation of one or more individuals as head- 
men, even though their actual power is circumscribed or even 
negligible. For convenience of exposition I shall call these officials 
"titular chiefs" in contrast to the "strong chiefs" possessing un- 
questioned authority. After discussing the functions of these two 
types of civil heads, I shall examine the factors that may have 
strengthened the titular chief's hands in the American milieu; and 
I shall likewise consider what agencies aside from chiefs of 
either category have assumed State functions. 

Titular chiefs. — Titular chiefs vary considerably in actual 
status. The Chipewyan individuals who bear the title exercise 
so little influence apart from the accident of personality that 
one might perhaps just as well put this north Canadian tribe into 
the chiefless category with the Eskimo and the Fuegians. Else- 
where the office is not only honorific, but also fraught with 
definite public functions. In order to overcome semantic difficulties 
it will be best to emphasize what the titular chief is not, before 
trying to indicate his positive attributes. That he cannot, in many 
American societies, correspond to an African chief is apparent 
whenever a single band or tribe has more than one title-bearer. 
Three hundred Canella are headed by three "chiefs"; another 
Ge people, the Pau d'Arco Kayapo, generally had two; the re- 
lated Gorotire band, five (in 1940). Until 1880 the Omaha had 
two principal chiefs, with a varying number of lesser ones; this 
oligarchy was then superseded by a septet of uniform rank. 
Among the Arapaho there were four chiefs, and the Cheyenne 
with a population never greatly exceeding 4,000 had forty-four! ^^ 
A series of examples from diverse culture areas will elucidate 
what American chiefs typically lacked. 

The Ojibwa (round Lake Superior) had a council "with vague 
and limited powers." It selected a chief "whose power was 
even vaguer than that of the council," and who was "less able 
to work his will against an existing custom." Tanner, who lived 

^' Birket-Smith, op. cit., p. 66; J. O. Dorscy, "Omaha Sociology," BAE, 3d Ann. 
Rept. (1884), p. 357; Curt Nimuendaju, "A Note on the Social Life of the Kayapo," 
American Anthropologist, XLV ( 1943 ) ; Llewellyn and Hoebel, op. cit., pp. 67 ff . 


Relation of Ethnology to Other Disciplines 

in this region from 1789 until 1822, mentions "the unstable power 
and influence of the chiefs." In an assembly of 1,400 Assiniboine, 
Cree, and Ojibwa, he remarks, "not one would acknowledge any 
authority superior to his own will." A chief was, indeed, entitled 
to some deference, "but this obedience . . . continues no longer 
than the will of the chief corresponds entirely with the inclina- 
tion of those he leads." About the same time the trader Tabeau 
notes that among the Teton Dakota "all authority is as naught be- 
fore the opposition of a single individual," and for the related 
Assiniboine, Denig — himself the husband of a woman of that 
tribe — offers an eye-witness's priceless corroboratory evidence. 
At a council attended by him the "leading chief" advocated peace 
with the Crow; a tribesman of lesser dignity vigorously and suc- 
cessfully opposed the idea, carrying the assembly with him. The 
historian Parkman, on the basis of personal experience in 1846, 
declares that very few Oglala Dakota "chiefs could venture with- 
out instant jeopardy of their lives to strike or lay hands upon 
the meanest of their people" and correctly notes the paradox 
that the "soldiers," i.e., police, "have full license to make use 
of these and similar acts of coercion." This institution will be 
discussed later. Among the Shoshoneans of Nevada, "any family 
was at liberty to pursue an independent course at any time"; 
in Arizona the head of the Maricopa had functions "more ad- 
monitory than coercive"; and among the Yuma the tribal leader, 
though appealed to in a dispute, was "more significant as an 
embodiment of spiritual power than as a lawgiver or executive." 
Equivalent testimony comes from Oregon and Washington.^^ 

"William Jones, "Central Algonkian," Annual Archeological Report, 1905 
(Toronto, 1906), p. 137; Edwin James, ed.. An Indian Captivity: John Tanner's 
Narrative of His Captivity among the Ottawa and Ofibwa Indians, Sutro Branch, 
California State Library, Occasional Papers, No. 20 (San Francisco, 1940), p. 151; 
Anne Heloise Abel, ed., Tabeau s Narrative of Loisel's Expedition to the Upper 
Missouri (Norman, Okla., 1939), pp. 105 f.; Edwin T. Denig, "Indian Tribes of the 
Upper Missouri," BAE, 46th Ann. Rept. (1930), pp. 430^56; Francis Parkman, 
Prairie and Rocky Mountain Life (Columbus, 1856), p. 291; Leslie Spier, Klamath 
Ethnography, Univ. Calif. Publ. Amer. Arch, and Ethn., XXX (1930), 33; idem, 
Yum-an Tribes of the Gila River (Chicago, 1933), p. 158; Ray, op. cit., p. Ill; 
Steward, op. cit., pp. 246-260; Grenville Goodwin, The Social Organization of the 
Western Apache (Chicago, 1942), pp. 178 f.; C. D. Forde, Ethnography of the 
Yuma Indians, Univ. Calif. Publ. Amer. Arch, and Ethn., XXVIII ( 1931 ), 134 f. 


Political Organization of American Aborigines 

Superficially the stratified societies of coastal British Columbia 
are different, but only superficially in the questions at issue. 
What they emphasize is social eminence, not political power. A 
Haisla chief "gives orders only in matters directly concerned 
with feasts and potlatches," — not in cases of quarrels, theft, or 
murder; the Tsimshian equivalent was responsible for his fol- 
lowers' safety in battle and indemnified the mourners if their kin- 
dred had been killed. How different from an African potentate 
who owns his subjects' bodies and collects all damages for in- 
juries sustained by them.^^ 

South America yields corresponding testimony. In British 
Guiana a Barama headman has limited authority. Each of the 
three Canella dignitaries works like everyone else; none of 
them wears a badge of higher status, or interferes in private 
affairs, or issues commands, or imposes penalties. Among the 
related Apinaye, the headman does initiate measures against a 
sorcerer, but he cannot order an execution without popular 
assent. To take two more Brazilian examples, Karaya villagers 
simply desert a chief whose actions they resent; and though a 
Nambikuara leader enjoys a good deal of influence, he "has 
no coercitive power at his disposal." In short, the typical Ameri- 
can chief may enjoy social standing, but he lacks sovereignty.^® 

What, then, are the titular chief's positive attributes and func- 
tions? The outstanding one forthwith explains the deficiency 
I have harped on: he refrains from attempting physical force, be- 
cause many societies conceive him as primarily a peacemaker. 
It would be a contradiction in terms for him to mete out punish- 
ment when his business is to smooth ruffled tempers, to persuade 

" Edward Sapir, "The Social Organization of the West Coast Tribes," Trans- 
actions of the Royal Society of Canada, Sec. II, 3d ser., Vol. IX ( 1915); R. L. Olson, 
The Social Organization of the Haisla of British Columbia, Univ. Calif., Antliro- 
pological Records, II (1940), 182; Franz Boas, "Tsimsliian Mythology," BAE, 
31st Ann. Kept. ( 1916), pp. 429 if., 499. 

"John Gillin, The Barama River Carihs of British Guiana, PMH, Papers, XIV, 
No. 2 (1936), 98, 140; Curt Nimuendaju, The Apinaye, pp. 19 f., 131 f.; idem. 
The Eastern Timhira, Univ. Calif. Pub. Amer. Arch, and Ethn., XLI (1946), 93, 
159-162, 239 f.; Fritz Krause, In den Wildnissen Brasiliens (Leipzig, 1911), p. 321; 
Claude Levi-Strauss, "The Social and Psychological Aspects of Chieftainship in a 
Primitive Tribe; The Nambikuara of North-Western Mato Grosso," Transactions 
of the New York Academy of Sciences, Ser. II, No. 1 ( 1944), p. 23. 


Relation of Ethnology to Other Disciplines 

the recalcitrant, coax and even bribe the justly aggrieved into 
forgoing vengeance. He is, indeed, a go-between of the Yurok or 
Ifugao order, but with the essential diflFerence of being the 
official, recognized, permanent moderator instead of a self- 
appointed one ad hoc. In order to compass his end — maintenance 
of communal harmony — he might stoop to eating humble-pie 
and to personal sacrifices. A Sanpoil chief presents each litigant 
with a blanket; his Cree colleague is expected to give up thoughts 
of revenge on his own behalf, such as other men freely indulge. 
A Winnebago went still further: "If necessary, the chief would 
mortify himself, and with skewers inserted in his back have 
himself led through the village to the home of the nearest kins- 
people of the murdered person." By thus arousing compassion 
he hoped to avert a feud.^" 

No wonder that an appeaser ex officio was not associated with 
warfare, was often — in his official capacity — deliberately divorced 
from violence and discipline. An Iroquois sachem's duties, Mor- 
gan reports, "were confined to the affairs of peace. He could not 
go out to war as a sachem." His position was sharply separated 
from the military leader's, being hereditary in the clan, whereas 
a successful captain gained a "chiefly" title of another category 
by personal bravery. This polarity was widespread. In a Fox 
Indian (Wisconsin) council, the Quiet and the War Chief were 
complementary figures, as are the Pueblo Town and War Chiefs 
— the former being prescriptively a man of peace who must not 
even go hunting, the latter a policeman who threatens punish- 
ment. The Omaha neither let a chief head a raid nor even allowed 
him to serve as a subordinate officer of one. Again, "a man who 
has often been on the warpath," say the Pawnee, "becomes 
imbued with the desire to take scalps and capture ponies and 
is no longer fit to be chief." A Winnebago cliief always belongs 
to one clan, a policeman to another.^^ 

^ David G. Mandelbaum, The Plains Cree, AMNH, Papers, XXXVII (1940), 
222; Paul Radin, "The Winnebago Tribe," BAE, 37th Ann. Rept. (1923), p. 209. 

^ Morgan, op. cit.. Part II, chaps, ii, iv, v; WiUiam Jones, Ethnography of the 
Fox Indians, BAE, Bull. 125 ( 1939), p. 82; Titiev, op. cit., pp. 59-68; Elsie Clews 
Parsons, Pueblo Indian Religion (Chicago, 1939), pp. 154 f.; J. O. Dorsey, "Omaha 
Sociology," BAE, 3d Ann. Rept. (1884), p. 217; G. A. Dorsey and J. R. Murie, 
Notes on Skidi Pawnee Society, Field Museum of Natural History, Anthropological 
Ser., XXVII (1940), 112 f. 


Folitical Organization of American Aborigines 

This dichotomy prevails even where a fusion of civil and 
military preeminence seems at first blush easily realized. In 
several South American tribes the "chief" did lead war parties, 
but whereas he became a virtual autocrat on a raid he relapsed 
into his usual impotence on his return. On this point early 
sources on the Kariri and the Tapuya (eastern Brazil) agree 
with recent ones on the Taulipang (south of the Roroima) and 
the Jivaro (Ecuador). One North American phenomenon is in- 
structive in this context. The Iroquois League found it desirable 
to create two generals "to direct the movements of the united 
bands," but these ofiBcials never aspired to a dictatorship. To 
quote Morgan, "the essential character of the government was 
not changed. . . . Among the Iroquois this oflBce never became 
influential." ^^ 

In short, the conceptions of civil and of military leadership 
were distinct in America. There was sporadic tyranny even in the 
democratic Northern Plains societies, but it sprang from in- 
dividual bullying, usually supported by a powerful body of kin 
or from putative supernatural sanction, not from the coup d'etat 
of a captain returning drunk with success and filled with the 
ambition of a despot. 

Besides being a skilful peacemaker, the ideal chief was a 
paragon of munificence. This may hold more often in North 
than in South America, but instances are not wanting in the 
south. Thus, a Nambikuara headman constantly shares with his 
tribesmen whatever surplus of goods he may have acquired: 
"Generosity is the quality . . . which is expected of a new 
chief." In the north, this demand is constant. In Alaska, where 
the Eskimo were affected by the ideology of their Indian neigh- 
bours, the title of "chief" automatically devolved on that Nuni- 
vak who entertained most lavishly at village feasts. A chief of 
the Tanaina Athabaskans (about Cook Inlet) feeds and clothes 
the destitute, provides for the households of men away on hunt- 

^Theodor Koch-Griinberg, Vom Roroima zum Orinoco (Stuttgart, 1923), III, 
94; Martin de Nantes, Relation succinte et sincere de la mission dti pere Martin de 
Nantes (Paris, 1706), p. 103; Thomaz Pompeu Sobrinho, "Os Tapuias do Nordeste 
e a monografia de Elias Herckman," Revista do Instituto do Ceard, XLVIII ( 1934), 
18; Rafael Karsten, Blood Revenge, War, and Victory Feasts among the Jibaro hi- 
dians of Eastern Ecuador, BAE, Bull. 79 (1923), pp. 7 f.; Morgan, op. cit., Part II, 
chap. V. 


Relation of Ethnology to Other Disciplines 

ing trips, adopts orphans, and even pays for shamanistic services 
that are beyond a poorer tribesman's means. The coastal tribes 
of British Columbia, notwithstanding their emphasis on hered- 
itary status, insisted that a headman should validate his claims 
by frequent distribution of property. In the Plains area, cliief- 
tainship and niggardliness were mutually exclusive. To quote 
Wissler, "no Blackfoot can aspire to be looked upon as a head- 
man unless he is able to entertain well, often invite others to 
his board, and make a practice of relieving the wants of his 
less fortunate band members." The Cheyenne or tlie Crow had 
identical standards of behaviour.^^ 

A third attribute of civil leadership is the gift of oratory, 
normally to be exercised on behalf of tribal harmony and the 
good old traditional ways. Speaking of the Sherente, Nimuendaju 

On many evenings ... I saw the chief assemble the village. Step- 
ping in front of the semi-circle . . . , he would impressively and 
vividly harangue the crowd for possibly an hovir. Usually he began 
circumstantially explaining the half-forgotten ceremonial of some 
festival. . . . There followed a lengtliy admonition ... to pre- 
serve ancient usage. In conclusion, he would urge all to live in 
peace and harmony. . . . 

The extinct Tupinamba of coastal Brazil regarded a species of 
falcon as the king of his zoological class: "ils se fondaient sur le 
fair que cet oiseau se levait de bon matin et haranguait les autres 
oiseaux, tout comme le chef de la hutte le faisait chaque jour, 
a I'aube, dans les villages tupinamba." In the Chaco the con- 
temporary Pilaga merely postpone oratory until nightfall: "Ce 
prurit d'eloquence est commun a tons les caciques et constitue 
. . . un des principaux attributs de leur dignite. . . . Le theme 
habituel de ces harangues est la paix, I'harmonie et I'honnetete, 
vertus recommandees a tous les gens de la tribu." In charac- 
teristic fashion a Chiriguano explained to Nordenskiold the ex- 

^ Levi-Strauss, op. cit., p. 24; Margaret Lantis, "The Social Culture of the 
Nunivak Eskimo," Transactions of the American Philosophical Society, n.s.. Vol. 
XXXV ( 1946), Part III, p. 248; Cornelius Osgood, The Ethnography of the Tanaina, 
Yale University Publications in Anthropology, No. 7 ( 1937), p. 132; Sapir, op. cit.; 
Clark Wissler, The Social Life of the Blackfoot Indians, AMNH, Anthropological 
Papers, VII ( 1911 ), 23; Llewellyn and Hoebel, op. cit., p. 79. 


Folitical Organization of American Aborigines 

istence of a female head of the tribe: her father had taught 
her to speak in pubhc. Thousands of miles to the north, in the 
Shoshone vernacular a headman figures as "the talker," which 
"designates his most important function." Maricopa and Apache 
chiefs, too, were matutinal lecturers; and among the Havasupai 
(Arizona) Spier says: "it might be said not that a chief is one 
who talks, but that one who talks is a chief." ^^ 

In my opinion, then, the most typical American chief is not 
a lawgiver, executive, or judge, but a pacifier, a benefactor of 
the poor, and a prolix Polonius. 

Strong chiefs. — But not all chiefs were only titular. A relatively 
small, but significant, number of societies had genuine rulers. 
It is best to begin with an unexceptionable example, the Inca 
state, the outstanding American sample of Drs. Fortes and Evans- 
Pritchard's category A — political systems with a well developed 
governmental apparatus. 

The Inca emperor, ruling by divine right, undoubtedly did 
control means of coercion. Through an elaborate "bureaucracy" 
he exacted tribute from his subjects and directed their labours, 
even their private lives. He did not scruple to transfer masses 
of the population from one province to another in the interests of 
the dynasty. What elsewhere in the New World were private 
wrongs here became offences against the Crown and called for 
summary official penalties. 

Emblematic of autocracy were the trappings of royalty other- 
wise conspicuously rare in America. The ruler wore and carried 
impressive regalia, travelled in a litter borne by special attendants, 
kept a large harem, and surrounded his court with an elaborate 
etiquette. His corpse was prepared for preservation in the palace, 
and his favourite wives together with a suitable retinue were 
strangled to accompany their master to the hereafter.-^ 

Concerning the Aztec chief the authorities yield contradictory 
and confusing evidence, but it seems clear that he did not con- 

^' Metraux, La Religion des Tiipinamba, p. 179; idem, "Etudes d'Ethnographie 
Toba-Pilaga," Anthwpos, XXXII (1937), 390; Erland Nordenskiold, Indianerlebcn 
(Leipzig, 1912), p. 229; Steward, op. cit., p. 247; Leslie Spier, Havasupai Ethnog- 
raphy, AMNH, Anthropological Papers, XXIX, Part 3 (1928), 237 f.; idem, Yuman 
Tribes of the Gila River (Chicago, 1933), p. 158; Goodwin, op. cit., pp. 165 f., 178. 

^ Rowe, loc. cit. 


Relation of Ethnology to Other Disciplines 

form to the Inca pattern. He was apparently not closely identified 
with the supreme deity; and, notwithstanding fixed succession 
within a lineage, he could be deposed. The hereditary halach- 
uinic of the Maya probably wielded greater power, claiming 
tribute as well as military service and periodically examining 
subordinate chiefs in order to weed out pretenders. Significantly, 
both he and the sacred war leader travelled in a litter, a symbol 
of exalted rank also attached to a Chibcha monarch, who re- 
sembled his Peruvian parallel in other respects. He, too, re- 
ceived tribute, kept a seraglio, hedged himself about with cere- 
monials, and was buried with several wives and slaves. When 
he expectorated, an attendant caught the spittle in an extended 
cloth — a form of flunkeyism hardly conceivable among the Crow 
or Cheyenne. ^^ 

It may be natural to find a full-blown political system among 
the materially advanced populations whose very numerical 
strength requires some central control if there is to be any 
solidarity. But, interestingly enough, the outlines of such a 
system appear also in the tiny states of the south-eastern culture 
area of North America. This anomaly has been recently stressed 
by Steward. Indeed, the Natchez sovereign came very close to 
the Inca conception of royalty. He claimed relationship with 
the solar deity, his kinsmen ranking as "Little Suns"; held power 
over life and death; travelled in a litter; and in death was fol- 
lowed by wives and servants, his bones being laid to rest in a 
temple near those of his predecessors. His subjects were obliged 
to keep at least four paces away from his person and would 
hail him "with genuflections and reverences." Elements of this 
complex, such as the litter, characterize the Timucua of Florida 
and the Chickasaw of Northern Mississippi; and though the 
monarchical principle is generally weaker in the south-east as 
a whole than among the Natchez, it reappears in full force in 
Virginia. "As halfe a God they esteeme him," Captain John 
Smith reports in writing of the Powhatan chief. This ruler 
arbitrarily ordered his subjects to be beaten, tortured, and killed, 
and kept a sizeable bodyguard to execute his will. "What he 

=«Roys, op. cit., pp. 192 f.; Tozzer, op. cit., pp. 165, 222; Kroeber, The Chibcha, 
p. 946; Vaillant, op. cit., pp. 113 ff. 


Political Organization of American Aborigines 

commandeth they dare not disobey in the least thing." He de- 
manded tribute of skins, beads, corn, and game; and numerous 
concubines waited upon him. Here and there undemocratic 
usages turn up as far north as New England, where they have 
been plausibly ascribed to south-eastern influences. However 
that be, the specific resemblances among Peruvians, Natchez, 
and Powhatan suggest a common origin for so atypical an Ameri- 
can polity.^' Of course, this does not imply that the social scheme 
diffused from the Inca Empire itself, a chronologically impossible 
assumption, but rather that certain elements of a monarchical 
system crystallized somewhere between Yucatan and Peru and 
spread in a period considerably antedating the expansion of Inca 
sovereignty. If I understand Professor Steward correctly, this 
agrees with his recent interpretation of the facts. 

Given the marked libertarian bias of most American aborigines, 
how can we conceive the growth of absolutism? What could 
convert the titular chief who cajoled liis tribesmen into preserv- 
ing the social equilibrium into a veritable king? 

Evolutionary germs. — In reexamining the chiefless or virtually 
chiefless tribes we discover here and there that the Indians 
willingly subordinate themselves to some individual for a par- 
ticular enterprise. In a rabbit drive the Washo and neighbouring 
Shoshoneans of the western Basin temporarily followed a leader 
noted for his skill as a hunter, though "apart from that special 
occasion his authority was nil." ^^ 

An exceptionally large gathering may favour the similarly 
spontaneous acceptance of a director. The Yahgan, who normally 
move about in very small groups, unite up to the number of 
eighty when a beached whale provides food for the participants 
at an initiation ceremony. Without an election some mature man 

^ J. R. Swanton, Indian Tribes of the Lower Mississippi Valley and Adjacent 
Coast of the Gulf of Mexico, BAE, Bull. 43 (1911), pp. 100-110, 139 ff.; idem. 
The Indians of the South-Western United States, pp. 161, 175, 598 ff., 641-654, 
728, 730; Regina Flannery, An Analysis of Coastal Algonquian Culture, The Catho- 
lic University of America, Anthropological Series, No. 7 (1939), pp. 116 f., 122 f.; 
J. H. Steward, "American Culture History in the Light of South America," South- 
western Journal of Anthropology, III (1947), 97. 

^R. H. Lowie, Notes on Shoshonean Ethnography, AMNH, Papers, XX (1924), 
196 f ., 284 f., 305. 


Relation of Ethnology to Other Disciplines 

well posted in traditional usage emerges as the master of cere- 
monies and henceforth plans the daily routine. What is more, 
he appoints a constable, who in turn chooses a number of 
deputies. These policemen exercise genuine legal authority: they 
forcibly drag refractory tyros to the initiation lodge, overpower 
a troublemaker, bind him, and let him lie for half a day without 
food or drink. The Yahgan, furthermore, have a men's club: the 
members as a whole bully their wives into fetching fuel and food 
for the assemblage, and one man has the duty of keeping women 
from prying.^^ 

Informally established offices are not necessarily ephemeral. 
The Nambikuara illustrate the rise of a relatively stable chief- 
taincy, as suggestively described by Levi-Strauss, A man with 
inborn gifts of leadership forms the nucleus for a group that 
voluntarily acclaims him, thereby shifting responsibility to liis 
shoulders. He directs the food quest during the difficult dry 
season, shares his surplus freely, prepares arrow poison for his 
adherents, and plans their entertainments. In requital, they con- 
cede him certain prerogatives, such as plural marriage, but with- 
out their approbation he is powerless. Here, then, there emerges 
a titular chief with genuine influence, though still not a ruler.^° 

By way of contrast there is a short-lived but absolute authority 
of the war leader as already noted for several South American 
groups. For North American parallels we have fuller data. A 
Crow supposedly organized his raid only when prompted by a 
supernatural patron, whence the leader's ascendancy over all 
who joined his expedition: theirs were the menial tasks, his the 
loot to dispose of as he chose, but also the responsibility for 
failure and losses. The equalitarian attitudes of everyday life 
recede, supplanted by a transitory overlordship. Omaha captains 
even appointed policemen who had the right to beat refractory 
or lagging warriors. Fleeting dictatorship of this limited range 
is not irrelevant to our problem. About 1820 the Cheyenne 
conceived themselves as one huge war-party, whose leader thus 
automatically became supreme, supplanting the tribal council of 

"■* Martin Gusinde, Die Feuerland Indianer, II: Die Yamana; vom Leben und 
Denken der Wassernomaden am Kap Hoorn (Modling bei Wien, 1937), pp. 199- 
208, 653, 779 fF., 798 ff., 805-961, 1319-1376. 

^ Levi-Strauss, op. cit., pp. 21 ff. 


Tolitical Organization of American Aborigines 

"chiefs." Yet in consonance with native ideology he retained not 
a vestige of his special authority when his task was done.^^ 

Undisputed supremacy for a restricted period was also granted 
during religious festivals. When a Hopi ceremony is in process, 
Stephen learned, "the chief of it is chief of the village and all 
the people." Similarly, the priest who directed a Crow Sun 
Dance was not merely the master of ceremonies, but the tem- 
porary ruler of the tribe, superseding the camp chief.^^ 

Non-chiefly authority. — Perhaps the most remarkable instances 
of authority, full-fledged and not altogether ephemeral, turn up 
in connection with important economic undertakings which are to 
be safeguarded in the common interest. 

A pertinent phenomenon from northern Brazil seems to have 
eluded general notice. The Apinaye chief, if properly qualified, 
succeeds his maternal uncle in the office, by virtue of which he 
guards the villagers' interests and orders the execution of evil 
sorcerers. But at the planting season a pair of men representing 
the moieties begin to act as independent executives. One of them 
collects the seeds, invokes the Sun to prosper them, and is the 
first to plant a plot. Both of these officials watch the crops, chant 
daily songs to promote growth, and forcibly prevent or punish 
premature harvesting. "Woe to any Indian woman who should 
dare to remove clandestinely even the most trifling product from 
her own plots before maturity is officially announced!" If the rule 
is broken, they "attack the houses of the villages or the camp, 
raging and throwing everything about pellmell, breaking the 
vessels and flogging with thorny whips any women who have not 
fled in good season, or gash them with a special weapon. . . ." 
Even the chief's wife was once severely chastized for transgress- 
ing the law. Apart from the religious feature, the phenomenon 
reminds an Americanist of the Winnebago or Menomini con- 
stables who punished overhasty gatherers of wild rice.^^ 

The last-mentioned officers from the Woodlands of North 

^* J. O. Dorsey, op. cit., p. 321; Llewellyn and Hoebel, op. cit., p. 163. 

^^ A. M. Stephen, Hopi Journal, ed. by Elsie Clews Parsons, Columbia University, 
Contributions to Anthropology, Vol. XXIII, Parts 1 and 2 (1936), 728. 

'"^ Nimuendaju, The Apinaye, pp. 13, 19, 89, 131 f.; Paul Radin, "The Winnebago 
Tribe," BAE, 37th Ann. Rept. ( 1923), pp. 226 f.; Alanson Skinner, Social Life and 
Ceremonial Bundles of the Menominee Indians, AMNH, Anthropological Papers, 
XIII (1913), 26. 


Relation of Ethnology to Other Disciplines 

America are obvious variants of the familiar Plains Indian "sol- 
diers" mentioned by Parkman. Their activities developed most 
spectacularly during a communal hunt, upon whose outcome the 
very life of the natives would depend. In order to ensure a maxi- 
mum kill, a police force — either coinciding with a military club, 
or appointed ad hoc, or serving by virtue of clan affiliation — issued 
orders and restrained the disobedient. In most of the tribes they 
not only confiscated game clandestinely procured, but whipped 
the offender, destroyed his property, and, in case of resistance, 
killed him. The very same organization which in a murder case 
would merely use moral suasion turned into an inexorable State 
agency during a buffalo drive. However, Hoebel and Provinse 
have shown that coercive measures extended considerably be- 
yond the hunt: the soldiers also forcibly restrained braves intent 
on starting war parties that were deemed inopportune by the 
chiefs; directed mass migrations; supervised the crowds at a major 
festival; and might otherwise maintain law and order. ^* 

Here, then, we find unequivocal authoritarianism. Theoreti- 
cally, the police acted, at least in a number of tribes, under the 
direction of the tribal chief or council. The foundation was thus 
laid for either an autocracy or an oligarchy. Why did this logical 
end fail to be consummated? 

In the first place, let us revert to the seasonal rhythm of the 
Plains Indians. During a large part of the year the tribe simply 
did not exist as such; and the families or minor unions of famihes 
that jointly sought a living required no special disciplinary or- 
ganization. The soldiers were thus a concomitant of numerically 
strong aggregations, hence functioned intermittently rather than 

^ Wissler, The Social Life of the Blaclcfoot Indians, pp. 22-26; idem, Societies 
and Ceremonial Associations in the Oglala Division of the Teton-Dakota, AMNH, 
Anthropological Papers, XI (1912), 17, 24; idem. The American Indian (New 
York, 1922), pp. 161, 178; Jane Richardson, Law and Status among the Kiowa 
Indians, American Ethnological Society, Monographs, I (1940), 9 f.; Diamond Jen- 
ness. The Sarcee Indians of Alberta, National Museum of Canada, Bull. 90, Anthro- 
pological Series, No. 23 (1938), pp. 11, 41; D. G. Mandelbaum, op. cit., pp. 203, 
205; A. L. Kroeber, Ethnology of the Gros Ventre, AMNH, Anthropological Papers, 
I ( 1908), 147 f.; E. A. Hoebel, "Associations and the State in the Plains," American 
Anthropologist, XXXVIII (1936), 433-438; idem, The Political Organization and 
Law-Ways of the Comanche Indians, AAA, Memoirs, No. 54 (1940), p. 82. 


Political Organization of American Aborigines 

Secondly, the "constitutional" relationship of chief and police 
was by no means so simple as might appear. It was definitely not 
that of the head of a modern state toward his army. Denig, whose 
observations on the impotence of Assiniboine chiefs have been 
quoted, ascribes to the police "the whole active power of gov- 
erning the camp or rather of carrying out the decrees and deci- 
sions of the councils." He himself witnessed "two killed and many 
severely thrashed for their misdemeanours." Were the soldiers, 
then, strictly subordinate to the council, as Denig's phraseology 
implies? Well, according to the same authority, if councillors 
threatened to grow violent at a meeting, "two soldiers advanced 
to the middle of the lodge and laid two swords crosswise on the 
ground, which signal immediately restored order and quiet." 
There was thus a dispersal of sovereignty: the titular chief had 
none, the council was in principle a governing board controlling a 
police squad that carried out their decisions, but de facto the 
theoretically subordinate police acted with considerable inde- 

The much fuller data on the Cheyenne collected by Messrs. 
Llewellyn and Hoebel corroborate this interpretation.^^ Here a 
self-perpetuating council of forty-four "chiefs" with safe tenure 
during a ten-year term of ojBfice was headed by five priest-chiefs, 
one of whom took precedence as the representative of the 
mythical culture hero, Sweet Medicine. This did not make him 
the equivalent of a Shilluk king, for he "wielded no consequent 
special political authority" nor was he above the traditional law. 
Unlike other Plains peoples, the Cheyenne for ritual reasons 
conceived homicide as a crime. When Little Wolf, the head chief 
and a man of superb record, killed a tribesman, though under 
mitigating circumstances, he did not escape the penalty, but went 
into voluntary exile. A lesser chief is known to have been severely 
flogged by the soldiers for a similar offence and was likewise 
banished, though not demoted in rank. 

To turn to the council as a whole, it is true that they ap- 
pointed one of the five existing military clubs to oversee a 
migration or a communal hunt. But, apart from such matters as 

E. T. Denig, op. cit., pp. 436, 439, 442, 444 f., 448, 455, 530 ff. 
Llewellyn and Hoebel, op. cit., pp. 67-131. 


Relation of Ethnology to Other Disciplines 

directing travel, the "chiefs" were httle concerned with secular 
affairs, sometimes waiving the right to a definitive decision and 
thus leaving a great deal to the discretion of their appointees. 
Accordingly, the police became the final authority in a large 
number of issues either beyond the competence of their electors 
or deliberately turned over to them by the council for settlement. 
The soldiers thus could, and repeatedly did, inaugurate legal 
precedents, nor does it appear that these were ever challenged by 
the "chiefs." 

Llewellyn and Hoebel draw attention to an extraordinary il- 
lustration of police autonomy. During a march directed by the 
Fox society, a councillor named Sleeping Rabbit answered a 
taunt by shooting the interlocutor, a member of the Dog organiza- 
tion. The arrow could not be extricated. The Foxes severely 
mauled and kicked the criminal; and when the victim's arm grew 
worse they decreed that Sleeping Rabbit must amputate it, a 
novel verdict. Public sentiment, crystallized in the four other 
societies, favoured exiling the culprit, but he avowed his guilt 
and, in self-infliction of a fine, presented the Foxes with five 
good horses. This settled the matter. 

As our authorities show, this was emphatically not an example 
of composition. Damages accrued neither to the victim nor to 
his kin nor to his society, but to the Foxes. They were the State 
in this case, receiving the indemnity as a Bantu ruler might in 
corresponding circumstances. Of course, so far as we know, the 
case is unique and might have remained so throughout Cheyenne 
history; but the mere possibility of its occurrence is significant. 

The relations of the Cheyenne council and soldiers were, of 
course, determined by the general American conception of 
chieftaincy. If more than temporary sovereignty were to be at- 
tained at all, it would thus more naturally centre in the police. 
Here we encounter a third factor that militated against autocracy 
or oligarchy. In this culture area the constabulary force was 
rarely fixed, being as a rule recruited differently for different 
seasons or even for specific occasions. In a Pawnee village, for 
example, the chief's adjutant and three of his deputies acted 
as police, but for a buffalo hunt a priest chose one of four 


Political Organization of American Aborigines 

societies as a nonce police.^^ The Cheyenne, we have noted, had 
five such organizations; it was not hkely that four of them would 
calmly submit to the oligarchical pretentions of one rival body. 

It so happens that in this tribe the Dogs did enjoy an unusual 
advantage over the other clubs: by an accident of history, a 
century or more ago, the males of one band collectively joined 
this society, so that in this solitary instance society and band 
coincided in adult male membership. The chief of the Dogs was 
thus ipso facto head of his band, and the Dog men remained 
united during the winter when rival clubs were scattered over 
various local divisions. Here, then, the germ for hegemony oc- 
curred, but it never reached fruition. 

A further point must be mentioned. Within any one of the 
military clubs its chief was supreme, issuing orders like a war 
captain and sometimes ruling his members with an iron hand. 
Yet the libertarian impulses of these Indians would not brook 
servility in an absolute sense; in 1863, characteristically, the 
Dogs forbade their chief to attend a treaty council with Ameri- 
can commissioners! 

In short, though the Plains Indians indubitably developed 
coercive agencies, the dispersal of authority and the seasonal 
disintegration of the tribes precluded a permanent State of 
modern type. Generalizing for the whole of America, there were 
sundry gropings towards centralization of power, but counter- 
acting trends made them fall short of permanent results. Yet such 
results were achieved in Peru and in so relatively simple a setting 
as that of the Powhatan. What were the circumstances involved 
in these cases? And is it possible to detect similar factors in the 
normally libertarian societies? 

The religious factor. — When Alexander the Great aspired to 
imperial grandeur, he was not content with tlie glory of a suc- 
cessful general, but claimed divinity and, as a mark of its ac- 
ceptance, prostration. This sacred character, we have seen, sup- 
posedly belonged to the Inca ruler and to the Natchez Great 
Sun; the obeisances and genuflections in their presence are the 
equivalent of Alexander's demand for proskunesis. With frankly 

^^ Dorsey and Murie, op. cit., p. 113. 


Relation of Ethnology to Other Disciplines 

evolutionary aim I shall assemble some data from the simpler 
American tribes in order to show that religious beliefs were used 
to attain political influence there; and I suggest that the awe 
which surrounded the protege of supernatural powers formed the 
psychological basis for more complex political developments. 
It is possible for a titular chief to add to his standing by combin- 
ing spiritual blessings with civil eminence, or he may enter an 
alliance with the religious functionary, thus foreshadowing the 
familiar spectacle of State and Church joined in the support of 
the established order. 

The latter contingency is classically exemplified in Gayton's 
study on the Yokuts,^^ a Central Californian stock of some 18,000 
souls divided into over fifty autonomous tribelets probably never 
exceeding 800 in population. In each of these units an acceptable 
member of the Eagle lineage served as chief, representing the 
mythical Eagle who had ruled the world in dim antiquity. 
Notwithstanding this lofty role, the chief was not an autocrat, 
but he did hold more than nominal precedence. Provided with 
food by his tribesmen, enjoying a monopoly of trade in highly 
prized products, entitled to a share in doctors' fees, he was the 
wealthiest man in the community. By way of reciprocity, it was 
his duty to entertain visitors, to help the poor, and to contribute 
generously to the cost of festivities. He determined movements 
from and to the village and alone could authorize the death 
penalty for a public enemy. In general, he adhered to the part 
of a peace-preserving headman, rarely making a vital decision 
without previously consulting other venerable men. 

Nevertheless, a chief could de facto magnify his power with 
the aid of a favourite shaman. In lieu of taxation the Yokuts 
expected the persons attending a festival to defray the expenses. 
If a wealthy villager evaded this obligation, the chiefs medicine- 
man would smite him with illness and impoverish his victim by 
exorbitant fees for sham treatment. Since the chief's consent 
was essential for violent measures against the doctor, he could 
always dismiss complaints on the subterfuge of insufiBcient evi- 
dence. It is important to note that public opinion as a rule 

^ A. H. Gayton, Yokuts-Mono Chiefs and Shamans, Univ. Calif. Publ. Amer. 
Arch, and Ethn., XXIV (1930). 


Folitical Organization of American Aborigines 

sympathized with the chief and the shaman, for the miser who 
failed to contribute at festivals thereby imposed extra burdens 
on his fellows. 

Given the native faith, an unscrupulous chief could evidently 
work his will in collusion with a shamanistic accomplice. Yet 
in the long run, Dr. Gayton explains, such knavish tricks led to a 
revulsion of feeling. A chief could not safely give rein to his 
malevolent inclinations. In the face of continuous suspicion his 
prestige would wane, in extreme instances he might even be sup- 
planted in office by a less objectionable scion of his line. As for 
his accessory, the attitude toward doctors being ambivalent 
here, as in much of North America, a persistently malevolent 
leech was likely to be killed by the enraged family of his victim. 
In short, the Yokuts system involved a considerable strengthen- 
ing of chiefly influence without, however, approaching anything 
like despotic rule. Its instructiveness lies largely in demonstrat- 
ing religion as a prop of the civil head on the relatively low 
plane of a simple hunting people. 

In a not inconsiderable number of South America societies 
there is a personal union of temporal and spiritual functions. 
In Colombia, the Kagaba and the Ijca (the latter linguistic 
relatives of the Chibcha ) do not dissociate the concepts of priest 
and chief. Among the Yaruro (Venezuela) each moiety rec- 
ognizes a shaman as its head. In tlie Mato Grosso the Tupi- 
Kawahib chief is "first of all, a shaman, usually a psychotic 
addicted to dreams, visions, trances and impersonations." An- 
other Brazilian group, the Botocudo, had as the leader of a 
band the "strongest" man, the epithet designating not muscular 
strength, but spiritual ascendancy. And, suggestively enough, 
these chiefs played a greater role than their colleagues in neigh- 
bouring populations and were in higher measure responsible for 
their bands, which sometimes took their names from the leaders'.^^ 

But even the Botocudo chief's influence pales before that of 

^'' Gustaf Bolinder, Die Indianer der tropischen Schneegebiete (Stuttgart, 1925), 
pp. Ill fF., 126 fF.; K. T. Preuss, "Forschungsreise zu den Kagaba-Indianern der 
Sierra Nevada de Santa Marta in Kolumbien," Anthropos, XIX-XV (1919-1920), 
364-368; Vincenzo PetruUo, The Yaruros of the Capanaparo River, BAE, Bull. 123 
(1939), p. 215; Niniuendaju, "Social Organization and Beliefs of the Botocudo of 
Eastern Brazil," pp. 97 ff.; Levi-Strauss, op. cit., p. 25. 


Relation of Ethnology to Other Disciplines 

the prophets who periodically arose in both Americas. ^° In my 
opinion, Nimuendajii, Spier, and Metraux have demonstrated that 
these messiahs did not so much react against white aggression, 
which represented merely a special case of the generic problem 
of evil, as against the supposed doom that threatened to engulf 
the moribund universe. Typically, the prophets promised salva- 
tion to their adherents, whereas unbelievers were to be trans- 
formed or destroyed. Given the mental atmosphere of the ab- 
origines, the more dynamic of the messiahs undoubtedly gained 
an extraordinary sway over their fellows. One of the early 
Guarani deliverers affected the pomp of royalty: refusing to 
walk, he had himself carried on the shoulders of his attendants; 
the common herd were not allowed to approach his person. Such 
pretensions rested on a claim to supernatural inspiration or to 
divinity itself. Many of the self-styled saviours tyrannically im- 
posed their will against common sense and, what is far more, 
against previously entrenched beliefs. In order to dance and 
chant as required, Obera's Guarani followers ceased to plant and 
harvest their crops in 1579. In the nineteenth century, under 
the spell of successive prophets, the Apapocuva band of this 
people repeatedly chased the will-o'-the-wisp of an earthly 
paradise, undertaking lengthy migrations to escape the menacing 
catastrophe. A little over a century ago an Algonkian messiah 
successfully ordered his people to kill their dogs and to abandon 
their hitherto prized sacred bags. For a while, about 1805, Ten- 
skwatawa, the Shawnee prophet, even held the power over the 
lives of his tribesmen, having his opponents burnt as witches. 

It is a far cry from the unstable sovereignty of these prophets 
to the close-knit Inca state, but the gap is far greater between 
the nominal chiefs described by Tanner among the Central 
Algonkians and the messiah he met in the very same tribe. The 
former were obeyed when the people so chose; at the latter's 

'"Mooney, op. cit., pp. 662, 672 fF., 676, 686, 700; Leslie Spier, The Prophet 
Dance of the Northwest and Its Derivatives: The Source of the Ghost Dance, 
General Series in Anthropology, No. 1 (Menasha, Wis., 1935); A. Metraux, "Las 
Hommes-dieux chez les Chiriguano et dans I'Amerique du Sud," Revista del Insti- 
tuto de Etnologia de la Universidad Nacional de Tucumdn, II (1931), 61-91; Curt 
Nimuendajii, "Die Sagen von der Erschaffung und Grundlagen der Religion der 
Apapocuva-Guarani," Zeitschrift fiir Ethrwlogie, XLVI (1914), 284-403. 


Tolitical Organization of American Aborigines 

behest they humbly killed their dogs, gave up their strike-a-lights 
at the expense of "much inconvenience and suffering," and threw 
away their hitherto holiest possessions. Assume the urge to 
leadership, as found by Levi-Strauss in the Mato Grosso, to be 
combined with an awe-inspiring supernatural sanction, and the 
way is clear to a formative stage on the way toward a govern- 
ment by divine right. What military prowess failed to create in 
aboriginal America is demonstrably possible even in a demo- 
cratic environment under the hypnosis of religious exaltation 
and the moral duress that follows in its wake. 

It is not part of my plan to squeeze out of the evidence con- 
clusions it will not bear. I cannot trace in detail the sequence of 
events that led from Ona "anarchy" to the close-knit structure 
known as the Inca state. I rest content with sketching a probable 
line of development. The totalitarian concentration of power in 
Inca Peru is an historic fact; so is the absence of any comparable 
official authority over most of the New World. If, for the sake 
of throwing the problem into relief, we assume an otherwise 
unwarranted teleological point of view, we discover sundry 
gropings towards the establishment of political authority, which, 
however, lose themselves in blind alleys. On analogy, what seems 
simpler than a military despotism under the two Iroquois gen- 
erals? Yet nothing of the sort arose in the face of an antagonistic 
cultural tradition. Similarly, the workings of the Cheyenne 
military societies seem to predestine the tribe to an oligarcliical 
system; but that, too, was precluded by the regnant pattern of 
social life. 

Nevertheless, equalitarianism recedes when confronted with 
putative supernatural favour. The very same men who flout 
the pretensions of a fellow-brave grovel before a darling of 
the gods, render him "implicit obedience and respect." It is 
probably no mere coincidence that Pontiac was a higher priest 
in the most sacred organization of his people, that Tecumseh 
was seconded by his brother, the prophet, and on occasion him- 
self laid claim to supernatural powers. The foundation of a major 
state, I suggest, was due to men of this type — men who both 


Relation of Ethnology to Other Disciplines 

imagined a unity beyond that of immediate kinship and con- 
tiguity and who simultaneously succeeded in investing their 
mission with the halo of supernaturalism. When not pitted 
against the terrible odds actually encountered by Pontiac and 
Tecumseh, natives of their mentality would be able to overcome 
both the dominant separatism and the dominant libertarianism 
of their fellows and create the semblance of a modern state. 


PART IV Theories and 

Characteristic or Lowie's writings on theoretical subjects is his 
appraisal of the work of particular men. To this approach his out- 
standing book The History of Ethnological Theory (1937) is ample 
testimony. It may interest the reader to compare the chapter on 
Tylor in that book with the article on him which Lowie wrote as 
an obituary for the American Anthropologist in 1917 (No. 25 in the 
present volume), or to compare the chapter on Lewis H. Morgan 
with "Lewis H. Morgan in Historical Perspective" (No. 26), written 
at almost the same time (1936) for A. L. Kroeber's Festschrift. 
Boas was so recurrent a preoccupation for Lowie that at least one 
of not less than seven papers on his great teacher should be in- 
cluded here (No. 29). Laufer's careful historical studies were much 
admired by Lowie and his students were grounded in the work of 
that distinguished sinologue. Although Laufer is less in fashion 
these days, an unpublished appreciation of him is included here 
(No. 32), as I believe Lowie would have wished. 

"A New Conception of Totemism" (No. 22) and "On the Prin- 
ciple of Convergence in Ethnology" (No. 23) are two of Lowie's 
early classics, and no collection of his writings would be complete 
without them. The first was stimulated by his attentive reading of 
Goldenweiser and the second by his equally thoughtful reading of 

Lowie was by no means a cultural antievolutionist, as a careful 

reading of even this incomplete series of papers will reveal. His 
interpretation of what constitutes cultural evolution, however, war- 
rants consideration. Therefore "Evolution in Cultural Anthropol- 
ogy: A Reply to Leslie White" (No. 28) has been included here, 
although I have on the whole avoided reproducing his more con- 
troversial articles as well as his more readily available ones. 

The two parallel articles, "Cultural Anthropology: A Science" 
(No. 27) and the hitherto unpublished "The Development of 
Ethnography as a Science" (No. 33) serve to illuminate Lowie's 
strongly rational and empiric convictions concerning science. So 
also does the article "Some Problems of Geographical Distribution" 
(No. 30). The reader may wish to supplement the articles with 
"Ethnography, Cultural and Social Anthropology" (American An- 
thropologist, LV [1953], 527-534). Finally, "Contemporary Trends 
in American Cultural Anthropology" (No. 31), which Lowie wrote 
two years before his death, is a characteristically fair, informed, 
and objective appraisal of the field to which he devoted his life. 


A New Conception of Totemism 

The significance of Dr. Goldenweiser's recent paper 
on totemism ^ lies in the fact that it presents for the first time 
what may be legitimately called "an American view of totemism," 
— "American" not only because it takes into account the data 
of American ethnography, but in the far more important sense 
that it is a view based on methodological principles which are 
becoming the common property of all the active younger Ameri- 
can students of ethnology. 

According to the traditional view, totemism is an integral 
phenomenon which is everywhere essentially alike. Thus, in 
Frazer's latest work on the subject, Totemism and Exogamy, the 
burden of proof is explicitly thrust on the shoulders of those 
who question the identity of totemic phenomena in dijfferent 
quarters of the globe and who uphold the theory of convergent 
evolution. In Part I of his paper "Australia and British Columbia," 
Dr. Goldenweiser has anticipated this challenge. He selects the 
series of features that are commonly regarded as distinctive of 
totemism, and compares the forms they assume in the two areas 

American Anthropologist, XIII (April-June, 1911), 189-207 
^ A. A. Goldenweiser, Totemism, an Analytical Study, reprinted from Journal of 
American Folk-Lore, XXIII (April-June, 1910), 115 fF. 


Theories and Theorists 

considered. The result is sufficiently striking. On superficial con- 
sideration, it appears that the Australian totem group resembles 
the clan of British Columbia in the exogamic regulation of mar- 
riage. But this resemblance is not significant; in both cases the 
exogamous character of the clan is not a primary, but a derivative 
trait. Because the clans are, in both areas, parts of the larger 
pliratric units, and because these phratries are exogamous, the 
totem clans must be exogamous, even though the clan, as a clan, 
may have nothing to do with exogamy. In other features, the 
totem clans of Australia and British Columbia clearly diverge. 
In Australia the social importance of the clan dwindles into 
insignificance as compared with that of the phratry; in British 
Columbia the clan is the social unit, par excellence. On the 
Northwest coast there is evidence for the development of the 
clans from village communities, such as nowhere exist in Aus- 
tralia. Finally, the American clans are graded as to rank — a 
condition likewise lacking in Australian totemism. In the matter 
of clan names, what similarity exists is again of a superficial kind. 
In Australia all clans are named from their totems; in British 
Columbia clans frequently derive their names from localities. 
But precisely where the American social divisions (phratries) 
are named after animals, we occasionally find that the eponymous 
animal is not identical with the crest animal, which is the one 
that corresponds, in religious function, to the Australian totem. 
If phratries are compared it is found that those of the Tlingit 
and Haida bear animal names, but that only a few of the 
Australian phratry names are definitely known to refer to animals. 
The view that the totemite is a lineal descendant of his totem 
is clearly developed in Australia; on the Northwest Coast, on 
the other hand, there is a fundamental belief in kiwian descent: 
the crest animal is one which has in some way been associated 
with the human ancestor of the group. Nevertheless, the author 
points out, there are myths in which the association is very close, 
and, in one group of traditions, the ancestor is the crest animal 
transformed. These instances, instead of militating against the 
author's point of view, constitute in reality strong evidence in 
support of it. For the myths in question result from the reaction 
of the guardian-spirit concept upon the basic belief that human 


A New Conception of Totemism 

beings have human ancestors. Now, the guardian-spirit concept 
is practically foreign to Australia. What similarity there exists 
between the Australian and the American myths is accordingly 
an ideal instance of convergent evolution. There remains the 
criterion of the taboo against eating or killing totems. Of this 
phase of totemic life Australia remains the classical example; 
in British Columbia, on the other hand, not a single instance of 
totemic taboos has hitherto been discovered, though there is 
an abundance of taboos of non-totemic character. 

A survey of the currently assumed symptoms of totemism in 
the two areas discussed thus reveals far-reaching diflFerences. It 
would be artificial, however, to confine the comparison within the 
limits set by conventional definitions of totemism. If we wish to 
disabuse ourselves of the preconceptions expressed in these 
definitions. Dr. Goldenweiser insists, we must not neglect to 
consider those cultural features which are empirically found in 
intimate association with the criteria generally recognized as 
totemic. In Australia, two elements have risen to so commanding 
a position within the totemic complex that each has been as- 
sumed as the essence and starting-point of totemism generically. 
These elements are the intichiuma ceremonies conducted for the 
multiplication of the totem animals, and the belief in the re- 
incarnation of ancestral spirits. On the Northwest Coast of 
America, analogous features are indeed found, but they are 
wholly dissociated from totemic institutions. A parallel condition 
of affairs is revealed in viewing the dominant traits of social life 
in northwestern America. The social life of the Kwakiutl is 
unintelligible without taking into account the groups of in- 
dividuals sharing the same guardian spirit; among the tribes 
farther north the clan tradition is essentially an account of the 
ancestor's acquisition of his guardian spirit, while the circum- 
stances incident thereto are dramatized in the dances of the 
secret societies. In Australia guardian spirits are rare, and, where 
found, are generally quite distinct from the totems; even when 
the two concepts do coincide, the guardian-spirit factor is of 
relatively slight moment. A second trait of special significance 
in the American area is the relationship of totemism to art, — 
the saturation of practically all decorative attempts with totemic 


Theories and Theorists 

motives, and the retroactive tendency to give, secondarily, a 
totemic interpretation to designs purely decorative in origin. This 
intimate connection is largely dependent on the quasi-realistic 
style characteristic of Northwest American art. In Australia, 
where geometrical motives predominate, art has exerted but little 
influence on totemic life. 

The empirical consideration of the totemic complex in the 
two typical regions dealt with thus establishes the essential 
diversity of the phenomena compared. The dominant motives of 
Australian totemism are not the dominant motives of Northwest 
American totemism, and vice versa. What resemblances exist 
are either superficial, or are functions of traits not directly as- 
sociated with totemism. Here, however, the criticism might be 
made that totemism in the areas selected is not comparable be- 
cause the American institution represents a far later stage of 
development. "The totem," as the author puts it in anticipation 
of this stricture, "has become attenuated to a crest, to a symbol; 
the living, flesh and blood relationship with the totem animal 
has been transferred into the realm of mythology; and, naturally 
enough, the taboo on the totem animal has dwindled away and 
finally disappeared." Dr. Goldenweiser's answer is unmistakably 
clear. "To a retort of that character, I would answer that we may 
safely assert that there is not one phase of human culture, so far 
represented in an evolutionary series of successive stages of 
development, where the succession given has been so amply 
justified by observation of historic fact as to be safely adopted 
as a principle of interpretation" (p. 22). The evidence from 
Northwest America must thus be admitted as coordinate with 
that from Australia. 

Having demonstrated the validity of the theory of convergent 
evolution for the totemic phenomena of Australia and the North 
Pacific Coast of America, the author proceeds in the second part 
of his paper, "The Totemic Complex," to consider, one by one, 
the traditional elements of totemism, and to determine their 
mutual relations on the basis of the widest possible ethnographic 

In the first section of Part II, Dr. Goldenweiser takes the im- 
portant step of divorcing exogamy from the concept of totemism. 


A New Conception of Totemism 

A discussion of the data on the natives of Assam, the Nandi of 
East Africa, several Plains Indian tribes, and the Samoan-Fijian 
culture province leads to the conclusion that "clan exogamy, 
although a usual concomitant of the other totemic features, is 
not a constant, hence not a necessary, concomitant of the latter; 
and again, where the other features are absent, exogamy may 
nevertheless occur" (p. 55). 

The pages immediately following embody what is probably 
the author's most important and original contribution to the 
subject, — a critique of the concept of exogamy itself. Not only 
may clans exist independently of exogamic rules, and exogamic 
rules independently of clans, but even where clan exogamy does 
exist the union of exogamy with the clan unit may be a secondary 
feature. In the case of the Kurnai, to be sure, special conditions 
seem to warrant Howitt's conclusions that marriage was originally 
regulated by non-localized totem clans, that paternal descent 
effected a localization of the totems, and that consequently, in 
recent times, locality has appeared as the regulating factor. 
But we should not by any means be justified in transferring 
this mode of reasoning to other areas where different conditions 
prevail. In British Columbia, for example, it seems probable that 
the clans at one time occupied separate villages. Now, whether 
the clans were exogamous as clans, or because they happened 
to coincide with local exogamic divisions, becomes an open 
question as soon as the existence of exogamy dissociated from 
clans has been established. The point comes out even more clearly 
where kinship exerts an influence on matrimonial regulations. 
Among the Todas a purely objective investigation reveals a 
number of exogamous clans, as well as a series of matrimonial 
rules based on kinship. But, subjectively considered, the matter 
is quite different. The Toda merely knows that certain relatives 
through the father and certain relatives through the mother — 
all of them included in the term piiliol — are not marriageable 
persons for him. Besides members of other clans, an individual's 
piiliol group also embraces all the people of his own clan, but 
this fact does not seem to have been noticed by Rivers' in- 
formants before he pointed it out to them. Similarly, among the 
Blackfoot, members of a band are forbidden to intermarry, not 


Theories and Theorists 

as band members, but because they are considered blood rel- 
atives. In a manuscript by Sternberg, which the author lays 
under contribution for additional facts, the Gilyak are described 
as a people with exogamous gentes. Now, where gentes as such 
are the exogamous units, two gentes suffice for the regulation 
of tribal marriages, the men of gens A marrying women of gens 
B, and vice versa. This reciprocal relationship is precisely what 
does not obtain in Gilyak society. If the men of gens A marry 
women of gens B, the men of gens B are ipso facto debarred 
from marrying women of gens A, the two gentes being regarded 
as gentes of "sons-in-law" and "fathers-in-law," respectively. The 
men of B must thus marry women of gens C, and so forth. It 
is, accordingly, obvious that the gens as such does not determine 

The thesis is thus established that "when the fact of a given 
social group not marrying within itself is ascertained, the in- 
formation acquired is but partially complete" (p. 59). The ex- 
ogamous nature of a group, as objectively observed, may indeed 
be a primary trait; but it may also be a derivative trait, — a 
necessary consequence of other regulations not linked with the 
group as such. Dr. Goldenweiser is thus emboldened to inquire, 
whether the exogamic character of Australian totem clans is a 
primary or a secondary characteristic. Taking up first the simpler 
form of social organization typified by the Dieri, viz., two ex- 
ogamous pliratries subdivided into smaller totem groups, he finds 
that a given clan can not be considered an exogamic unit be- 
cause in no case are the exogamic marriage regulations fully 
determined by clan affiliation. If this condition did obtain, 
nothing would prevent members of clan a from marrying mem- 
bers of clan b of the same phratry. But this is emphatically not 
what takes place. In reality, clan a "behaves exactly as would 
an individual of phratry A if there were no clans" (p. 60). Be- 
cause it forms part of a larger unit exogamic in its own right, 
clan a must ipso facto be exogamic. The condition of affairs is 
strictly parallel to that among the Tlingit and Haida Indians. 
In Australia the derivative character of clan exogamy is il- 
lustrated among the anomalous Aranda (Arunta), where some 
clans occur in both phratries. Here a man of clan a may marry 


A New Conception of Totemism 

a woman of clan a if she belongs to a different phratry, but he 
must not marry her if she belongs to the same phratry. 

The argument is greatly strengthened by a consideration of 
the four-class system. For here the class is the marriage-regulat- 
ing unit, and the clans are in no sense exogamous units, as each 
clan contains two sets of members with distinct matrimonial 
regulations. In tribes with the eight-sub-class system a parallel 
argument holds: the sub-class is the marriage-regulating unit, 
and each clan consists of "four matrimonially heterogeneous 
units." The question arises, what, in these systems, may be the 
marriage-regulating functions of the phratry, and of the phratry 
and class, respectively? Dr. Goldenweiser is of opinion that in 
the four-class system the fact beyond doubt is the exogamy of 
the class, while that of the phratry remains to be investigated; 
in the eight-sub-class system the immediate data indicate the 
exogamous nature of the sub-class, and the matrimonial functions 
of the class and phratry remain to be investigated. The point to 
be determined would be the native feeling with regard to these 
larger units, — whether, for example, the phratry of four-class 
tribes continues in the minds of the natives to constitute a dis- 
tinct exogamic group. The necessity of taking into account the 
subjective attitude of the natives is strikingly illustrated in per- 
haps the most suggestive passage of the entire paper. A purely 
objective description of the regulations found among four-class 
tribes does not by any means necessitate the current mode of 
representation. Instead of subdividing two phratries into two 
exogamous classes, it is possible to unite the intermarrying classes 
into endogamous moieties with exogamous subdivisions. This has 
actually been done by Professor Klaatsch, a relatively naive 
observer in matters ethnological, in the description of Niol-Niol 
social organization. This traveler has even recorded native names 
for the endogamous moieties. Dr. Goldenweiser rightly insists 
that, objectively, Klaatsch's mode of representation is as legitimate 
as the one ordinarily employed. His suspicions are aroused merely 
by the fact that endogamous moieties recognized as such by the 
aborigines have hitherto escaped the eyes of other Australian 
ethnographers. Moreover, the class names of the Niol-Niol ap- 
parently correspond to those of the Aranda, whose phratry- 


Theories and Theorists 

class organization seems firmly established. Accordingly, the 
author does not contend that Klaatsch's scheme represents the 
subjective facts, though he admits that "on a par with the domi- 
nant phratric organization there may also exist in these Aus- 
tralian tribes a consciousness of the objectively endogamous 
groups constituted by the pairs of intermarrying classes" (p. 64 ) . 
In this extraordinarily illuminating discussion the critic can 
find fault only with the author's use of the term "exogamy." A 
word is obviously required to designate the rule against members 
of a group marrying among themselves, — in other words, the 
rule of the incest group. The word sanctioned by usage is the 
etymologically unexceptionable and self-explanatory term "ex- 
ogamy." Dr. Goldenweiser, however, modeling his conception of 
the "typical exogamous relation" on the conditions supposedly 
found in Australian tribes with two phratries, writes: "An exog- 
amous relation is fully represented only when both the group 
within which marriage is prohibited, and the one into which it 
is permitted or prescribed, are given" (p. 60). Accordingly, he 
views the class (among four-class tribes) as the exogamous unit 
par excellence, and finds an approach to "pure totemic exogamy" 
in the Arabana institution of each totem clan being permitted 
to intermarry with only one particular clan of the complementary 
phratry. Were the matter one purely of nomenclature, the re- 
definition of a current term would, of course, be perfectly 
legitimate. In the present instance, however, it seems to the 
critic that the term is not, and can not conveniently be, used with 
consistency in the modified sense. Where there are only two 
social units exogamous in their own right, intermarriage follows 
as a physical necessity; the group into which marriage is per- 
mitted or prescribed is determined by the mere statement of 
the prohibitory regulations. This is obviously not the case when 
there are four, or six, or fourteen groups, within each of which 
marriage is prohibited. To be sure, it might be said that in such 
instances the exogamous relation, in Dr. Goldenweiser's sense, 
is fully represented, inasmuch as, where statements to the con- 
trary are lacking, a member of group I may marry members of 
all other groups. But if positive regulations are to be taken into 
account, it certainly is not the same thing whether a man must 


A New Conception of Totemism 

marry into the only group existing besides his own, whether he 
must marry into one of a number of other groups, or whether 
he may marry into any of the other groups extant. To the 
critic it seems that there are only two alternatives. Either we 
adopt the author's conception of exogamy. Then the mutual 
relationship of intermanying classes with rules against intra-class 
marriage would form the standard illustration of exogamy; 
phratries would formally, but, for reasons just given, might only 
formally, exemplify exogamy; and it would be inadmissible to 
speak glibly of four exogamous Tsimshian clans (p. 9), of a 
great number of exogamous Khasi clans (p. 53), of fourteen 
exogamous Bahima clans and forty-one exogamous septs (p. 74). 
Or, we cling to the accepted usage of the term. Then exogamy 
may be ascribed to any group prohibiting marriage among its 
members. In this case, the exogamy of the Kamilaroi class, as 
well as the exogamy of the Arabana clan, is a derivative feature, 
— a logical consequence of phratric exogamy. In addition to this 
derivatively (and therefore relatively unimportant) exogamic 
trait, the Kamilaroi class and the Arabana clan have certain 
positive marriage-regulating functions, which, however, have 
nothing to do with exogamy, of which the functions are only 

In the next part of the section on "Exogamy and Endogamy," 
the author briefly mentions the constant tendency to extend 
regulations of marriage, even where fairly definite regulations 
already exist. An unusually suggestive instance is furnished by 
the Toda (p. 168). Within the (endogamous) Teivaliol moiety 
there are a number of exogamous clans. But the members of 
the Kundr clan outnumber the other clans to such an extent 
that the exogamous rule can only be followed by the Kundr 
marrying most of the members of the other clans. Thus very few 
of the latter are left to marry one another, and the condition of 
affairs seems to approach as a limit the widespread division of 
a tribe into two exogamous intermarrying phratries. The oc- 
currence of positive obligations for certain classes to intermarry 
— a point too little noticed by other writers — is strikingly il- 
lustrated by the Gilyak groups of prospective husbands and wives. 
Dr. Goldenweiser, in discussing the matrimonial institutions of 


Theories and Theorists 

this people, also calls attention to the correlated rules of "psychic 
intercourse." There is restriction of conversation and intimacy 
between persons who might come into conflict from jealousy, 
and avoidance obtains, in different degrees of stringency, be- 
tween relatives debarred from intermarrying. On the other 
hand, there is great freedom between prospective husbands and 
wives, and an extraordinary cordiality characterizes the rela- 
tions of fathers-in-law and sons-in-law. There can be little doubt 
that the correlation of the rules for sexual and psychic inter- 
course, which the author considers only in connection with a 
single tribe, merits more extensive investigation. 

The next two sections, "Totemic Names" and "Descent from 
the Totem," add little to the argument of the corresponding 
divisions of Part I. Some additional examples are adduced to 
show that eponymous totems, while remarkably frequent, do 
not occur universally, and that the totem is not invariably re- 
garded as the ancestor of the group. Under the heading "Taboo," 
the author points out that quite generally restrictions of conduct 
are associated with groups other than totem clans, while there 
are tribes, such as those of the Iroquois confederacy, whose 
totem clans are not connected with taboos against killing or 
eating the eponymous animal. In reply to the specious reasoning 
of many writers, that totemites abstain from killing or eating 
their totems because they regard them as kin. Dr. Goldenweiser 
intimates that taboos may have a variety of origins. The Omaha 
furnish a telling series of illustrations. For here many of the 
taboos associated with totem groups are logically unconnected 
with the totems, and it seems practically certain that each of 
these "fanciful prohibitions" had a distinct origin. 

In the pages on "The Religious Aspect of Totemism," the 
author emphasizes the fact that totemism and animal cult are 
distinct phenomena. Animal worship is prominent where totemic 
groups do not exist. On the other hand, worship of the totem is 
very rare, and in some cases there is a complete absence of 
religious associations with totems. It is obvious that under the 
circumstances it would be absurd to regard totemism as a form 
of religion, or as a distinct stage in the evolution of religious 
beliefs. On this point, at least, Dr. Goldenweiser finds himself 


A New Conception of Totemism 

in agreement with the views put forward in Frazer's most recent 
pubHcation on the subject. 

The comprehensive survey of ethnological phenomena in Part 
II thus confirms the conclusions arrived at from a comparison of 
Australian and British Columbian conditions. Totemism can 
no longer be considered as an integral phenomenon, Totemic 
complexes are "conglomerates of essentially independent features" 
(p. 88). It may be possible to trace logically the development of 
the several traits from a single hypothetical factor of fundamental 
importance, but only through historical proofs can such deduc- 
tions gain scientific value even for limited areas. Neither a 
system of naming groups after totem animals, nor the doctrine of 
descent from the totems, nor a religious regard for the totem, 
in fact, not one of the symptoms ordinarily assumed, is a con- 
stant feature of totemism; and there is no evidence for the 
historical or psychological primacy of any one of them. The 
instances of other factors — magical ceremonies in Australia, es- 
thetic motives in British Columbia — rising to prominence within 
the totemic complex illustrate the variability of the phenomenon 
studied, and lead to the important queries, "If totemism includes, 
roughly speaking, everything, is totemism itself anything in par- 
ticular? Is there anything specific in this phenomenon, or has 
the name 'totemism' simply been applied to one set of features 
here, to another set there, and still elsewhere perhaps to both 
sets combined?" (p. 89). 

Dr. Goldenweiser replies that, in the light of his foregoing 
analysis, the specific trait of totemism cannot be a certain 
definite sum of elements, but only the relation obtaining between 
the elements (p. 92). In a given totemic complex, factors a, b, 
c, etc., are associated and correlated so as to form a relatively 
integral combination. The fairly complete integration of totemic 
factors results from the fact that elements in themselves socially 
indifferent become associated with clearly defined social groups, 
the association being effected by means of descent (p. 93), In 
defining the relationship of the totemic elements, the author 
starts from a consideration of the current view that totemism has 
a religious and a social aspect. The occasional absence of any 
religious factor, notably among the Iroquois, induces him to 


Theories and Theorists 

eliminate the term "religious" and to conceive totemism as the 
association of "objects and symbols of emotional value" with 
definite social units, the latter being defined as units perpetuated 
through descent. Again, totemism is usually described as a static 
phenomenon. Yet, nothing is more obvious than its variability in 
time. Dr. Goldenweiser's investigation, accordingly, culminates 
in the dynamic definition: "Totemism is the process of specific 
socialization of objects and symbols of emotional value" (p. 97). 

As an epilogue, the substance of which would have more ap- 
propriately preceded the definition of totemism, comes a dis- 
cussion of "Origins, in Theory and History." Schmidt's, Frazer's, 
and Lang's theories are jointly subjected to a methodological 
critique. Instead of attempting to understand present conditions 
on the basis of their established antecedents, these theories 
select a prominent feature of modem totemism and project it 
into the past, assuming it to be the starting-point of the totemic 
process. This, the author contends, is unjustifiable; for what is 
now of overshadowing significance need not always have figured 
with equal conspicuousness. The second step made by the 
theorists mentioned, namely, the deduction of other features 
from the one assumed to be primary, is likewise illegitimate; 
for it assumes the unity of the totemic features and a uniform 
law of development. The former assumption has been refuted 
by the preceding analysis, while the latter seems doubtful in 
the light of modern research. Finally, the authors criticized 
err in neglecting the influence of borrowing on the development 
of culture in a given area. 

To bring home this last point. Dr. Goldenweiser proceeds to 
show what the course of totemic development has actually been 
in the carefully studied region of British Columbia. While the 
southern Shuswap have the loose village organization typical of 
the Salish tribes of the interior, the western Shuswap have a 
social system obviously patterned on that of the coastal tribes, 
and indirectly derived from them. Among other instances within 
the same area, the transformation of the institutions of certain 
Athapascan tribes is especially remarkable. Such features as 
potlatches, clan exogamy, and an hereditary nobility, have 
been obviously borrowed from neighboring coastal tribes; and 


A New Conception of Totemism 

in so far as the Athapascan tribes possessing these traits differ 
in the details of these institutions, the differences can some- 
times be directly explained by contact with correspondingly 
differing tribes of the coast. The actual history of such changes 
could never be foretold by means of speculations as to primitive 
psychology; it was ascertained only by intensive study of the 
influences to which each tribe has been subjected (p. 109). 
In the data already accumulated on Australia, Dr. Goldenweiser 
finds evidence of the far-reaching influence of diffusion on 
cultural development; and his paper terminates in the con- 
fident prophecy that future research will reveal conditions of 
borrowing comparable to those established in British Columbia. 
As English ethnologists seem to adopt only with reluctance 
the historical point of view advocated by other students, it may 
be well to recall Tylor's memorable words: 

Most of its phenomena (that is, of human culture) have grown 
into shape out of such a comphcation of events, that the laborious 
piecing together of their previous history is the only safe way of 
studying them. It is easy to see how far a theologian or a lawyer 
would go wrong who should throw history aside, and attempt to 
explain, on abstract principles, the existence of the Protestant 
Church or the Code Napoleon. A Romanesque or an Early English 
cathedral is not to be studied as though all that the architect had 
to do was to take stone and mortar and set up a building for a given 
purpose {Researches into the Early History of Mankind, p. 4). 

The historical significance of Dr. Goldenweiser's essay will 
perhaps become clearer from a parallel between the develop- 
ment of ethnological thinking and the evolution of philosophical 
thought in general. Popular philosophy has always had the tend- 
ency to assume a necessary bond between the constituents of 
a relatively stable complex of observed elements, — to assume 
that there is a "thing" which has properties, an ego which has 
sensations, feelings, and other manifestations of consciousness. 
Valuable as such summaries of experience are from a practical 
point of view, they become indefensible from a higher standpoint. 
The ideas we form of "things" result from an association (by 
contiguity) of the ideas of its properties. A child learns by ex- 
perience that a brown patch of color and a certain form of 


Theories and Theorists 

resistance to the touch are Hnked together, and by connections 
of these ideas develops the idea of a table. A "thing" is thus 
nothing distinct from its properties; it is nothing but the sum- 
total of these properties; there is no mystic unity in reality apart 
from the properties. (Cf. Hoffding, Psychologie, pp. 212, 226, 
285.) The ethnologist, like the uncritical philosopher, is con- 
fronted at every step with conjunctions of features which at first 
seem indissolubly united. A geometrical pattern is associated in 
the primitive craftsman's mind with some definite animal or 
plant. It is natural to assume that the association is a primary 
one, — that the design is a degenerate attempt at realistic rep- 
resentation. Games are played as means of divination or processes 
of sympathetic magic. Should they not be conceived as cere- 
monial contrivances? Tales of heroic exploits culminate in the 
hero's ascension to the sky. Must not the whole plot be a function 
of his celestial affiliations? Social units with animal names and 
food taboos prohibit marriage within the group. To regard names, 
taboos, and exogamous rules as merely manifestations of the 
same fundamental phenomenon is, at a relatively early stage of 
inquiry, the obvious and psychologically most intelligible thing 
to do. 

At a more critical stage, however, the instability of the com- 
plexes attracts notice. What was at first supposed to be a neces- 
sary connection is reduced to a mere conjunction of elements. 
Thought is no longer arrested by a contemplation of the mystic 
underlying units and their relations with the observed elements; 
to determine the nature and inten'elations of these elements 
themselves becomes the highest, nay only possible, goal of in- 
vestigation. In the domain of physical science, a critical reforma- 
tion of this type has been, within recent decades, effected by 
Professor Ernst Mach. In ethnology, the school which has set 
itself a corresponding aim, which endeavors to supplant the 
traditional belief in mystic ethnological complexes with a deeper, 
though, it may be, still only proximate, analysis into provisional 
elements, is the school headed by Professor Franz Boas. Under 
his influence Kroeber and Wissler have shown that the same 
pattern is subject to varying interpretations even within the 
same tribe: design and interpretation are found to correspond 


A New Conception of Totemism 

to distinct psychological processes. An analogous conclusion with 
regard to the conjunction of story plot and cosmic phenomena 
has been drawn by the present writer. Independently of Boas, 
but in thorough harmony with his point of view, Seler, in criticiz- 
ing Preuss, and Haddon, in criticizing Culin, have pointed out 
en passant that the association of ritual with forms of diversion 
is a secondary development. What all these writers have at- 
tempted in the study of their own problems Dr. Goldenweiser 
has done for the far more complicated subject of totemism. He 
has shown the futility of attempting to connect any definite con- 
ception of concrete ethnological facts with the term "totemism." 
He has shown that there is no justification for assuming a com- 
mon substratum underlying all the "totemic" complexes: a com- 
plete statement of all the social, religious, esthetic, and other 
correlates with their interrelations, as found in a given area, 
exhausts the possibilities of description and explanation. 

However, as already shown in the resume of the section on 
"The Complex in the Making," Dr. Goldenweiser does not aban- 
don the term "totemism," but seeks to justify its retention by 
a redefinition of the word from a dynamic standpoint. It is here 
that he passes beyond the limits reached by his fellow-students 
of secondary associations. For, while the latter are generally 
content to indicate the fact that a secondary association of ele- 
ments has occurred. Dr. Goldenweiser boldly undertakes to 
define, with some precision, the process itself of the association. 
That is to say, he does not merely hold that totemism is the 
result of a secondary association of social units with various 
factors. He holds, in addition, that the association resulted from 
the fact that objects and symbols which were originally of 
emotional value only to individuals became, through descent, 
values for definite social groups (p. 97). 

Before entering into a critique of this conception, it is worth 
noting that many forms of association not ordinarily considered 
totemic would be classed as such according to the new definition. 
A phratry and a local group might illustrate the dynamic process 
in question as well as any "totem kin" of other writers. In par- 
ticular, the fact that the name occupies no favored position, but 
appears as but one factor of many that may be associated, seems 


Theories and Theorists 

to render "totemism" almost all-inclusive. This is especially the 
case when we consider that, on the author's theory, it is not at 
all necessary that the names be derived from animals or plants. 
Discussing Iroquois totemism (p. 96, footnote), Dr. Golden- 
weiser argues that even here, where the totem is merely a name, 
it, at least formerly, represented an emotional value, inasmuch 
as otherwise the name would not have become firmly fixed in 
social groups. Obviously, the same reasoning — which the reviewer 
cannot consider conclusive — would apply to local units with 
non-animal names. It is not clear whether, or where, the author 
would draw the line here; indeed, the data bearing on names 
of totem groups require more extensve treatment than that given 
in the present paper before it will be possible to form a clear 
view of Dr. Goldenweiser's conception of this special point. 

Dr. Goldenweiser's definition of totemism may be considered 
from two points of view. In how far does it accurately represent 
the phenomena commonly designated as totemic? And, to what 
extent does it represent the totality of phenomena which seem 
psychologically and sociologically related with these totemic 

In reply to the first query, it must be admitted that the author's 
definition outlines a plausible course of development. Neverthe- 
less, it is possible to conceive that conditions other than those 
defined by Dr. Goldenweiser may lead to typical totemism. As- 
sume two locally distinct groups, each with its own taboos against 
the eating of a certain animal. Then the union of these two 
groups would lead to a typical totemic society, in the ordinary 
sense of the term, if we add the feature of exogamy. Such a 
hypothetical development in no way militates against the author's 
general point of view. Nevertheless, it is perfectly easy to un- 
derstand the process, from what we know of the development 
of taboos, without recourse to the theory that the taboo was 
originally of only individual significance and afterwards be- 
came socialized through descent. Or, to take a case which is 
not hypothetical. What evidence is there to show that among 
the Iroquois the clan name was originally an individual posses- 
sion which, through descent, became socialized? To exclude in- 
stances of this type from the list of totemic phenomena by a 


A New Conception of Totemism 

rigorous application of the definition would reduce the whole 
discussion to a logomachy, which would be entirely beside the 
author's purpose. For what he attempts to do is precisely to 
define the essential features of the process resulting in what 
are ordinarily called totemic phenomena. The fundamental ob- 
jection to such a definition as Dr. Goldenweiser has attempted is 
that it is frequently impossible to determine whether it correctly 
represents the historical process of association. If we assume the 
association of name and social group as the starting-point of 
totemism — and, as the author himself has shown, this combina- 
tion sometimes exhausts the content of totemism — it is, in our 
ignorance of the actual history of the development, impossible 
either to prove or to refute the theory that the group names, 
not only in the Iroquois, but in the Australian cases as well, 
ever served to designate individuals. The inherent probability 
of such a condition does not seem very great. If the association 
of taboo and social group is taken as the starting point, the a 
priori probability of a socializing process will presumably appear 
considerably greater to the majority of ethnologists. Nevertheless, 
the hypothetical instance given above seems to indicate that 
socialization is not a Denknotwendigkeit for the comprehension 
of the established association. The critic is therefore of opinion 
that a non-committal attitude on the process of association (so 
far as it eludes observation) is highly advisable. Totemism would 
then be defined, not as a socialization of various elements of 
(at least potentially) emotional value, but merely as the associa- 
tion of such elements with social groups. 

The second question is, does Dr. Goldenweiser's conception 
embrace all the phenomena essentially related to those of totemic 
phenomena generally recognized as such? The writer feels that, 
inclusive as is Dr. Goldenweiser's definition, it limits the field 
of totemism too narrowly by an exaggerated emphasis of the 
element of descent. By a "complete social unit" Dr. Goldenweiser 
understands one group of at least two within the tribe, each in- 
cluding both men and women, and perpetuated by descent (pp. 
93, 94, 97, 98). Accordingly, in dealing with the resemblance be- 
tween totemic institutions and religious societies whose members 
share the same guardian spirit, he does not discover a genuine 


Theories and Theorists 

homology. "While a certain psychological affinity between the 
two institutions is not improbable, their genetic relationship, 
claimed by some, calls for demonstration" (p. 94). The matter 
of genetic relationship may be dismissed at once as irrelevant, 
for as Dr. Goldenweiser, on the very next page, states his belief 
in the convergent evolution of totemic phenomena, absence of 
genetic connections would not, from his point of view, bar re- 
ligious organizations from the fold of totemic institutions. Their 
exclusion, then, rests essentially on the definition of a social unit. 
Now, the definition given by Dr. Goldenweiser seems to the 
writer quite arbitrary. If the peculiarity of totemic phenomena 
lies only in the relation obtaining between the elements (p. 92), 
the psychological resemblance of this relationship would seem 
to be the predominant issue, while the precise nature of the 
social group becomes negligible. Among the Gros Ventre (At- 
sina), where every man passes successively through a series of 
age-societies, these grades are well-defined social units. The 
association with each of them of a certain animal for which 
several of the societies are named does not seem to differ in 
principle from the association of a clan with its crest or epony- 
mous animal ancestor. It may not be out of place here to refer 
to the fact that Schurtz has already darkly hinted at a con- 
nection between totemism and the age-grades of the northern 
Plains Indians { Alter sklassen und Mdnnerhiinde, p. 154). The 
argument just advanced in behalf of age-societies is obviously 
applicable to the type of religious societies specifically mentioned 
by Dr. Goldenweiser, as well as to still other forms of social units. 
Is their exclusion justifiable from a point of view that emphasizes 
merely the relation of elements entering into a "totemic" complex? 
In advancing these comments, the writer is fully aware of the 
fact that he may not have fully grasped Dr. Goldenweiser's mean- 
ing. The subject of totemism is not yet quite in the position of 
those metaphysical problems of which Clifford has said that, 
in discussing them, people find it peculiarly difficult not only 
to make out what another man means, but even what they mean 
themselves. But that it is peculiarly difficult to discover another 
man's conception of totemism is amply attested by the recent 
history of ethnology. However this may be. Dr. Goldenweiser 


A New Conception of Totemism 

himself knows quite well that his analytical study is not definitive, 
but programmatic; that the next step must be a more extensive 
ethnographic investigation of the field. What he has already 
given is a statement of first principles. Whatever deficiencies 
may be found in his definition, he has been the first to show 
at length, and with irrefragable logic, that totemism can not be 
treated as an integral datum, — the first, as already stated, to 
apply the doctrine of secondary association to the subject of his 
inquiry. From this point of view, his paper constitutes a land- 
mark in the history of totemic study, — the prolegomena to all 
positive attempts at a sane interpretation of "totemic" institutions. 



On the Principle of Convergence 
in Ethnology 


In a recent work on the methods of ethnology,^ Dr. 
Graebner once more expounds the theoretical position familiar 
to readers of his former writings.^ The central problem of eth- 
nology is for liim the determination of cultural connections. 
Resemblances in culture must be primarily accounted for by 
historical connection, — in the first place, because the existence 
of such connection stands unchallenged for a large part of the 
phenomena; secondly, because there are no objective criteria of 
independent development. Lack of historical relationship can- 
not be established by the most intense feeling that such a relation- 
ship is improbable, for this feeling is of a purely subjective 
character. Neither can the absence of proof for historical con- 
nection be interpreted as a stringent demonstration that an 

Presented at the annual meeting of the American Folk-Lore Society in Washing- 
ton, December 28, 1911, and printed in Journal of American Folk-Lore, XXV 
(1912), 24-42. 

^ Fritz Graebner, Methode der Ethnologie (Heidelberg, 1911). 

" More particularly, idem, "Die melanesische Bogenkultur imd ihre Verwandten," 
Anthropos, IV ( 1909 ), 726-780, 998-1032. 


Principle of Convergence in Ethnology 

historical relationship does not exist. It is indeed conceivable 
that, after determining all cultural relationships, we may still 
be confronted with independent partial similarities; but obviously 
this conclusion would result, not from the application of definite 
criteria of independent evolution, but solely from the non- 
applicability of the criteria of cultural connection. "So bleibt 
denn als erstes und Grundproblem der Ethnologic wie der 
ganzen Kulturgeschichte die Herausarbeitung der Kulturbezie- 

What, then, are the criteria of cultural connection? Two such 
are recognized by Graebner, — the criterion of form, that is, of 
the coincidence of characteristics not necessarily resulting from 
the nature of the objects compared: and the criterion of quanti- 
tative coincidence. In innumerable cases the form-criterion is 
self-sufficient. Nevertheless, Graebner notes instances of its mis- 
application, through fanciful affiliation of heterogeneous forms. 
Here, it seems, the quantitative criterion should have been used; 
that is to say, as it is one of the cardinal doctrines of Graebner's 
philosophy of ethnology that the diffusion of isolated cultural 
elements — even of myths — is impossible ( kulturgeschichtliches 
Nonsens), the doubtful parallelism of two forms can be im- 
mediately established if they are recognized as elements of the 
same or related cultural complexes. So far as continuous areas are 
concerned, these criteria have not been challenged: they are 
generally employed in establishing linguistic relationship, and 
have proved valid in the study of European culture. Graebner 
sees no reason for limiting the criteria to continuous areas: he 
does not hesitate, for example, to use them as proofs for a far- 
reaching connection between Old World and New World culture. 
The only objection advanced against such applications of the 
criteria has been the improbability, under primitive conditions, 
of diffusion over the tremendous distances dealt with. On the 
one hand, this argument is refuted by the migrations of the 
Malay o-Polynesians and the occurrence of Asiatic tales in South 
America. But, in addition, the contrary argument may be strength- 
ened by two auxiliary principles. The supposed lack of con- 
tinuity between two areas may prove deceptive. There may be 
found cultural features bridging the geographical gap between 


Theories and Theorists 

the areas compared ( continuity-criterion ) ; and there may be such 
a diffusion of cultural elements, that geographical proximity 
varies directly with the degree of cultural relationship ( criterion 
of form-variation), — a result manifestly not to be expected on 
the theory of independent evolution of parallel forms. ^ 

The foregoing account aheady describes by implication Graeb- 
ner's position on the subject of convergent evolution. From 
his point of view, it matters little whether similarities are be- 
lieved to result from a psychology common to mankind or from 
the convergence of originally distinct phenomena. In either 
case, there is an assumption of independent development; and 
as positive criteria of independent development are, according 
to Graebner, non-existent, both theories are on a methodologically 
inferior plane as compared with the doctrine of historical connec- 
tion. In particular, Graebner criticises Ehrenreich's definition 
of "convergent evolution" as the result of similar environment, 
similar psychology, and similar cultural conditions. Similari- 
ties in natural conditions, he contends, have been considerably 
overestimated. The psychology of different branches of man- 
kind shows as much differentiation as their physical traits. As 
a matter of fact, the psychological unity of mankind, which is 
invoked to explain cultural resemblances, has really been in- 
ferred only from the observed resemblances. If peoples of 
distinct geographical areas reveal far-reaching psychical resem- 
blances, the question arises whether these are not ultimately due 
to genetic relationship or cultural contact. So far as the similarity 
of cultural conditions is concerned, Graebner insists that, if in- 
dependent development be assumed, similarity of cultural con- 
ditions could result solely from the natural environment, and 
that similarity of cultural conditions would presuppose a high 
degree of psychical resemblance. Against Ehrenreich's state- 
ment that in spite of various parallels with Old World culture 
the culture of America bears a distinctively American stamp, 
Graebner declares that it is not clear how heterogeneous cultural 
conditions could lead to parallels, which, according to Ehrenreich, 
must be due to similar cultural environment. An a fortiori 
argument is used to clinch the discussion. European civilization 

* Graebner, Methode der Ethnologie, pp. 94-125. 


Principle of Convergence in Ethnology 

has developed a remarkable similarity of cultural milieu. Never- 
theless the number of well-authenticated instances of independ- 
ent parallel development is exceedingly small. In the majority 
of instances we find merely combinations of thoughts and motives 
already extant in the culture common to authors, inventors, 
or thinkers. But even the residual cases lose their force as to 
convergent development among primitive races: for, on the 
one hand, these modern instances rest on a peculiarity of modern 
culture, — the conscious striving for progressive development; 
on the other, the same thought may indeed be conceived twice, 
but the literature of science indicates that the same thought does 
not necessarily become socially and culturally significant in 
more than one case. If a cultural similarity resting on close 
genetic relationship has produced so small a number of independ- 
ent parallels of social significance, it may reasonably be doubted 
whether the relative psychological unity of mankind, and the 
resemblance of natural conditions, could produce such absolute 
identity of culture as to result not merely in the conception, 
but in the social acceptance and further development, of the 
same thoughts. 

Two questions confront the reader in connection with the 
views presented above. In the first place, does Dr. Graebner 
correctly define the logical standing of the antagonistic theories 
of independent development and genetic or cultural relation- 
ship? Secondly, does Dr. Graebner grasp the essentials of the 
doctrine of convergence as it has been employed in ethnological 
practice? The following pages will be devoted to an examination 
of these questions. 

The supposed methodological superiority of the theory of con- 
tact and relationship rests, as indicated above, on the assumption 
that it is distinguished by positive, objective criteria, while the 
rival theory lacks such criteria.^ Indeed, the argument that in- 
dependently evolved cultural similarities could be detected 
only by the non-applicability of Graebner's criteria (p. 107) 

* This point of view also appears in Graebner's brief reply to a critique by Arthur 
Haberlandt in Petermanns Mitteilungen, 1911, pp. 228-230. 


Theories and Theorists 

involves the strongest conviction that criteria of independent 
development not only have not been found, but that it is im- 
possible to discover them. 

In the first place, the objectivity of Graebner's criteria is in 
> large measure illusory. He himself points out that the form- 
criterion is liable to fanciful subjective interpretations (p. 118). 
In all doubtful cases, however, he counsels testing by the second, 
unconditionally objective {unhedingt ohjektiven) criterion of 
quantity. It may at once be admitted that this criterion does 
provide a quantitative measure for the degree of relationship 
between two cultural complexes. This relationship, however, 
cannot be established except by demonstrating the relationship 
of corresponding elements in the two complexes. Each equation 
can be made only by the application of the form-criterion. In 
each particular comparison there will thus admittedly be a 
subjective factor, hence it is quite illogical to argue that a sum- 
mation of parallels will eliminate the subjective element. Apart 
from this, what we know of the psychology of investigation 
does not justify us in the belief that a student who discovers 
intensive morphological resemblances — though other investigators 
fail to note them — would ever feel the necessity of resorting 
to a test by another criterion; and, if he did, he doubtless would 
have little dijBBculty in propping up his fanciful parallel by 
others not less whimsical. Indeed, the quantitative test leads to 
curious results in Graebner's own case. Against Haberlandt, — 
who reproaches him with classifying together such diverse ob- 
jects as the "male" and the "female" spear-thrower, nay, even 
the Maori sling-stick, — Graebner urges that, if a complex has 
once been established on the basis of well-defined elements, even 
a morphologically indeterminate element, such as the spear- 
thrower, must be regarded as part of the complex, provided 
its distribution coincide with that of the other elements.^ This 
is undoubtedly a vicious principle. From the identity of even 
an indefinitely large number of corresponding elements in two 
series it does not follow that certain other associated elements 
are genuine parallels and must be brought into a genetic relation- 

"^ Ihid., p. 229. Graebner, of course, does not neglect the differences in spear- 
thrower types except in his theoretical speculations (see Anthropos, IV, 736). 


Principle of Convergence in Ethnology 

ship. The "male" and the "female" spear-thrower might reason- 
ably be grouped together as conceivable differentiations from 
a common prototype; but to argue that so heterogeneous an 
object as the sling-stick is related to them if it occurs in a similar 
combination of elements is not testing the criterion of form, but 
sacrificing it. 

While Graebner's criteria of genetic relationship are thus found 
to lack the strictly objective character claimed for them, in- 
dependent development need not be defended on purely sub- 
jective grounds, even where a stringent demonstration is im- 
possible. Graebner criticises Ehrenreich for holding that the same 
mythological ideas may develop independently a great number 
of times from universally observable natural phenomena.^ This, 
he contends, is an a priori position lacking in sanity, because 
from the ready conceivability of independent development we 
cannot infer the fact of independent development (p. 97); that 
is to say, Graebner considers the theory of independent develop- 
ment inferior, because it leaves the door open to the arbitrary 
individual judgment of psychological probability. Now, it may 
at once be admitted that no amount of psychological investiga- 
tion can actually demonstrate that two given cultural phenomena, 
possessing as they do the unique character distinctive of historical 
happenings, originated independently. A demonstration could 
be given only if we knew the actual history, which we generally 
do not. As a matter of fact, however, the theory of independent 
development is not one whit worse oflF in this respect than its 
rival theory; for it is an utterly mistaken notion that the psycho- 
logical factor is excluded by the assumption of cultural relations. 
The comparison of form can never do more than establish the 
identity of forms; that such identity is to be explained by a 
genetic relationship is an hypothesis of varying degrees of 
probability. That the details of the crutch-shaped Melanesian 
paddle should occur in South America is to Dr. Graebner a 
sufiBcient proof of common origin (p. 145). Why? Because he 
cannot conceive how such similarity could result independently. 
But what is inconceivable for him is perfectly conceivable for 
Ehrenreich and others. From the inconceivability of independ- 

" Paul Ehrenreich, Die allgemeine Mythologie und ihre ethnologischen Grund- 
lagen (Leipzig, 1910), p. 266 


Theories and Theorists 

ent development by a single student we certainly cannot infer 
the fact of a common origin. We are dealing with probabilities, 
not with certainties in either case; the only point is to increase 
the probability of either theory, and here I cannot find that the 
doctrine of independent development is in a less favorable posi- 
tion. It seems to me, on the contrary, that a number of observa- 
tions in individual psychology, as well as a number of social 
facts, well-nigh establish the independent development of certain 
simple cultural traits; and that in other cases the probability of 
such development, while not as yet determined, can be readily 
investigated at the present time. 

As an example of the former kind I should regard certain 
observations on the reactions of children in the dark. If the 
widespread fear of the dark which enters into primitive beliefs 
were exclusively the result of tradition, it might be reasonably 
argued that it had developed from the same source of origin. 
This theory, however, becomes improbable as soon as we find 
that the distinctive feeling of uncanniness appears in equal force 
where all traditional beliefs tending to foster dread of the dark 
have been rigorously excluded from the child's curriculum. ''^ An 
element not altogether negligible in primitive belief is thus 
shown to be an element of our psycho-physical constitution. The 
psychology of dreams furnishes additional material bearing on 
the question. If certain physiological conditions, say retinal ir- 
ritations, are regularly correlated with certain dream images 
which coincide with widespread mythological conceptions, then 
such conditions must be considered as constituting a vera causa 
for the explanation of the mythological ideas. Thus, the wide- 
spread conception of a grotesquely distorted countenance may 
be plausibly traced to Wundt's "Fratzentrdume." Of course, we 
do not know, and never shall be able to know with certainty, 
that these dreams formed the foundation of the corresponding 
beliefs. But to disregard them entirely, to deny that they affect 
the merits of the case, would be to indulge in that form of 
sterile hypercriticism with which Graebner not infrequently 
reproaches his own opponents. In other directions, systematic 

''Ernst Macli, Die Analyse der Empfindungen (Jena, 1906), p. 62. These obser- 
vations were confirmed by Dr. Petrunkevitch in an oral communication to me. 


Principle of Convergence in Ethnology 

observations could at least be planned and instituted. For ex- 
ample, psychological child-study might establish the fact that 
children of different countries react in an essentially similar way 
on the everyday phenomena observable in the heavens. With 
the same reservations as before, due to the unique character 
of historical happenings, we should then be justified in attach- 
ing a high degree of probability to Ehrenreich's conjecture as 
to the independent origin of simple nature myths. In other fields, 
the study of individual psychology from this point of view 
might present greater practical difiBculties: it might, for ex- 
ample, prove impossible to disentangle the influence of traditional 
art-forms in an inquiry into the development of drawing and 
design. On the other hand, the inquiry into types of association, 
such as Galton was the first to conduct on a large scale, seems 
full of promise, especially so far as color and number symbolism 
are concerned. The contention that an apparently very odd as- 
sociation common to two distinct regions must have travelled 
from one to the other must immediately lose its force if we find 
the same association arising with a certain frequency among our- 
selves. The objection might indeed be raised that, in order to be- 
come a cultural phenomenon, the individual association would 
have to be socialized; this would, however, apply in equal meas- 
ure on the supposition of borrowing. 

So far, then, as the objectivity of the criteria is concerned, the 
inferiority of the theory of independent development stands un- 
proved. In determining genetic relationship on the ground of 
formal resemblance, the influence of the personal equation is 
unavoidable; on the other hand, the arbitrariness of speculations 
on independent development can be limited by the results of 
scientific (as opposed to popular) psychology. 

If there is any difference in the value of the two theories, it 
must rest on the alleged absence of historical proofs for inde- 
pendent development, in the face of the universally admitted 
existence of such proofs for historical connection. It remains to be 
shown that this allegation is erroneous, that there exist unex- 
ceptionable instances of convergent evolution. For this purpose 
it is necessary to examine somewhat more closely the concept of 



Theories and Theorists 


The fundamental error in Graebner's critique of convergent evolu- 
tion lies in the fact that it entirely ignores the group of phenomena 
to which the principle criticised has been most succesfuUy ap- 
plied. Taking into account only Ehrenreich's definitions of "con- 
vergence," and disregarding completely Ehrenreich's further re- 
marks on the subject, Graebner is led to reject the theory because, 
for the explanation of identities, it seems to involve the assumption 
of a mystic psychological unity (p. 145). 

To be sure, it must be admitted that, if we found exact paral- 
lels of very complicated phenomena, their occurrence in two 
areas, no matter how widely separated, could not reasonably be 
explained by convergence. Let us assume for a moment that we 
found on the northwest coast of America a social system duplicat- 
ing such Australian elements as four-class exogamy, belief in 
lineal descent from the totem, elaborate rites for the multiplica- 
tion of totems, and the like. If this were the fact, an explanation 
by the psychic unity of mankind would be lamentably deficient, 
as may readily be shown by examination of a concrete case. Ehren- 
reich writes, "Wo gleiche Geistesanlage sich vereint mit Gleich- 
heit der Wirtschaftsform und der gesellschaftlichen Stufe, wird 
die Cultur im Allgemeinen iiberall einen gleichen Charakter, einen 
gleichen Typus tragen, und wir diirfen uns nicht wimdern, wenn 
solche gleiche Typen audi in Einzelheiten grosse tjbereinstim- 
mung zeigen und Convergenzen hervorbringen." ^ Let us test the 
explanatory value of the principle, as thus defined, by a single 
example, Ehrenreich finds a surprising resemblance between the 
Dukduk masks of New Britain and the Fish-Dance masks of the 
Karaya, as well as between the correlated usages. Granting the 
resemblance, nay, even the exact identity, of the features in ques- 
tion, what meaning can we associate with the statement that the 
parallel is due to psychic resemblance linked with like economic 
and sociological conditions? The identity to be explained is not 
found except among the two above-mentioned representatives 
of two distinct racial types. What are the psychic traits and cul- 

^ Paul Ehrenreich, "Zur Frage der Beurtlieilung und Bevverthung ethnogra- 
phischer Analogien," Correspondenz-Blatt der deutschen Gesellschaft fiir Anthro- 
pologic, Ethnologie und Urgeschichte, 1903, pp. 176-180. 


Principle of Convergence in Ethnology 

tural conditions common to these two tribes, which are not shared 
by those of their geographical neighbors and racial congeners 
lacking the cultural homologies under discussion? The principle 
of continuity is in fact not less essential to a sane theory of in- 
dependent development than to a sane theory of transmission. 
There is at least no logical diflBculty in assuming that certain laws 
of evolution are immanent in human society, and must lead every- 
where to the same results. But to say that psychic affinity and 
cultural similarity have produced in two or in a few instances 
the same result is logically admissible only if it be shown at the 
same time for what specific reasons the same result is not notice- 
able in all other cases, even where psychic affinity is reinforced 
by racial relationship, and cultural affinity by geographical and 
historical contact. So far, then, as Graebner's attack is directed 
against Ehrenreich's explanation of supposed identities, it is en- 
tirely justified: such an explanation is indeed nothing but a mys- 
tification. Granted the existence of identities, they are inexph- 

But the entire aspect of the question changes if we do not inter- 
pret the given parallels as identical or homologous, but merely 
as analogous. In the brief but profound paper quoted above, 
Ehrenreich has treated this problem with the greatest possible 
clearness. Over and above what he regards as genuine conver- 
gences, he distinguishes "false analogies," due to the inadequacy 
of our knowledge, to the premature classification of diverse traits 
under the same concept, labelled with the same catchword. It is 
merely necessary to conceive all parallels of any degree of com- 
plexity as "false analogies," — to explain them as Ehrenreich him- 
self explains, in exemplary manner, the various forms of totemism, 
of the belief in metempsychosis, of the swastika and eye-omament, 
— and the mystical element in the theory of convergence disap- 
pears. The observation of similarities, especially in the absence of 
obvious paths of diffusion, then leads directly to the query whether 
the similarities are not purely classfficatory, and hence, from the 
standpoint of genetic relationship, illusory. 

In a review of Graebner's recent book, which has been pub- 
lished since the writing of the preceding paragraphs, Professor 
Boas says, 


Theories and Theorists 

Nobody claims that convergence means an absolute identity of phe- 
nomena derived from heterogeneous sources; but we think we have 
ample proof to show that the most diverse ethnic phenomena, when 
subject to similar psychical conditions, or when referring to similar 
activities, will give similar results (not equal results), which we 
group naturally under the same category when viewed, not from 
an historical standpoint, but from that of psychology, technology, 
or other similar standpoints. The problem of convergence lies in 
the correct interpretation of the significance of ethnic phenomena 
that are apparently identical, but in many respects distinct; and 
also in the tendency of distinct phenomena to become psychologi- 
cally similar, due to the shifting of some of their concomitant ele- 
ments — as when the reason for a taboo shifts from the ground of 
religious avoidance to that of mere custom.^ 

As is shown by a preceding quotation from Ehrenreich, Professor 
Boas goes too far in his initial statement, for Ehrenreich's con- 
ception of genuine convergence does practically involve a belief 
in an absolute identity derived from heterogeneous sources; but 
his utterance indicates that in America, at all events, convergence 
has been treated in a manner which entirely escapes Graebner's 

It is now necessary to discuss convergence as resulting from 
modes of classification, to show what form of classification gives 
rise to the appearance of identical results from diverse sources, 
and to illustrate the point by a number of special instances. 

Premature classification appears in ethnological literature in two 
principal forms: the ethnologist may either infer from the un- 
doubted identity of certain elements in two different complexes 
that the complexes themselves are identical; or he may fancy 
identity of elements or complexes where none exists. The first type 
of premature classification has wrought considerable mischief 
in the consideration of ceremonial complexes, such as the Mide- 
wiwin and the Sun Dance. The psychology of this fallacy is not 
unlike that of illusions. A complex such as the Midewiwin is 
described for some particular tribe; and some conspicuous feature, 
say, the shooting-ritual, acquires a symbolic function; so that 

® Franz Boas' review of Graebner, Methode der Ethnologie, in Science, XXXIV 
( 1911 ), 804-810; the quotation is from p. 807. 


Principle of Convergence in Ethnology 

whenever this feature appears in another tribe, it is at once sup- 
posed to indicate the presence of the residual elements of the 
complex first described. This would indeed be a justifiable in- 
ference, if a complex invariably represented a quasi-organic 
unit; but this is precisely what is not ordinarily the case. For 
example, Dr. Radin has recently shown ^^ that the Midewiwin of 
the Winnebago and that of the Central Algonkian are not identi- 
cal, because in each there has been a secondary association be- 
tween the common elements and a preponderant group of specific 
elements, which in large measure can be shown to result from 
the specific character of Central Algonkian and Winnebago 
culture respectively. I have suggested elsewhere ^^ that what 
Dr. Radin has successfully demonstrated for the Midewiwin ap- 
plies in like measure to the Sun Dance of the Plains tribes. We 
cannot reduce to a common prototype the various forms in which 
the ceremonies grouped under this catch-word appear. All we 
can do is to ascertain the relatively few common elements which 
have acquired the symbolic function mentioned, and to investi- 
gate their varying combinations in different cases. 

It is clear that the form of erroneous classification treated above, 
however large it may loom in ethnological discussion, has noth- 
ing to do with convergent evolution; for in the cases mentioned 
the genetic relationship of the identical features has never been 
challenged, while apart from these features there is obvious 
divergence. It is Ehrenreich's group of "false analogies" that sup- 
plies us with illustrations of the second type of classificatory error, 
and this has a direct bearing on the principle of convergence. 

Comparing the two types of inadequate classification, we may 
say that the first type involves the assumption that an organic re- 
lationship exists where it does not exist, while the second type 
of error results from the failure to note that the supposedly paral- 
lel elements are organically related to two distinct complexes. In 
this latter case, then, the parallelism is between logical abstrac- 
tions rather than between psychological and ethnological realities. 
Some concrete illustrations will make the matter clearer. 

^^ Paul Radin, "The Ritual and Significance of the Winnebago Medicine Dance," 
Journal of American Folk-Lore, XXIV (1911), 149-208. 

" R. H. Lowie, The Assiniboine, AMNH, Anthropological Papers, IV ( 1909 ), 77 ff. 


Theories and Theorists 

Owing to their theoretical interest, the so-called age-societies 
of the Plains may properly serve to introduce the subject, J. O. 
Dorsey reports that among the Omaha there were three feasting 
societies, composed of old men, middle-aged men, and youths 
respectively. In tribes of the same cultural area (Arapaho, 
Blackfoot, Mandan, Hidatsa) other writers have found series of 
dancing societies evincing a more refined classification by age, 
admission into any one society being contingent on a payment. 
Schurtz assumes that the existence of age-grades among the 
Omaha and other Plains tribes is due to an innate tendency of 
human society towards an age-grouping, which leads everywhere 
to similar results. From Graebner's point of view, the existence of 
so marked a feature as age-grades in a practically continuous area 
must be explained as due to historical connection. If, on the other 
hand, we here applied the principle of convergence in the sense 
defined by Ehrenreich, we should say that the resemblance be- 
tween the Omaha age-classes and the age-societies of the other 
Plains tribes is due to the union of general psychic and specific 
cultural similarities of all the tribes concerned. 

As a matter of fact, each of these three interpretations is er- 
roneous. The Omaha feasting organizations are age-classes prop- 
erly so called; that is to say, a man belongs to one of the three 
classes by virtue of his age. But the fact that, say, the Hidatsa 
societies present the appearance of age-classes is due to the mode 
of purchase obtaining in this tribe. The age factor is indeed active, 
inasmuch as it is customary for age-mates to purchase a society in 
a body; but there is no established division of Hidatsa society 
into age-grades, no correlation between age and membership in 
a certain definite organization. The correlation is, instead, be- 
tween membership and purchase: a Hidatsa belongs to every 
society of the series which he has purchased, but which has never 
been purchased of him. A man of ninety may thus hold member- 
ship in a young men's society, and under abnormal circum- 
stances a group of men may acquire a membership which ranks 
superior to that of an older age-group. To call both the Omaha 
and the Hidatsa organizations "age-societies" is therefore admis- 
sible only if we regard this term as a convenient catch-word which 
may denote neither psychologically nor genetically related phe- 


Principle of Convergence in Ethnology 

nomena. The age-factor that we isolate in studying the Hidatsa 
system is, of course, as a logical abstraction comparable to cor- 
responding abstractions, whether derived from the Omaha sys- 
tem or that of the Masai. In reality, however, it forms part of 
a context which determines it, and from which it cannot be 
wrested without completely altering its character. What we find 
in comparing the Omaha and the Hidatsa systems is therefore a 
convergence of a type different from that defined by Ehrenreich, 
but coinciding absolutely with that of his "false analogies," which 
result from our relative ignorance of the phenomena compared. 
So long as we knew only that the Hidatsa had societies composed 
of men of different ages, it was possible to classify them as age- 
grades proper. With the additional knowledge of the subjective 
attitude of the natives towards these societies, the justification 
for such a classification disappears. 

What has just been shown for age-grades may be similarly 
shown for the much-discussed phenomenon labelled "exogamy." 
It has commonly been assumed that the regulation against mar- 
riage within a certain group, no matter in what part of the globe 
such a regulation may be found, is uniformly the same in prin- 
ciple. Dr. Goldenweiser has recently shown that this is by no 
means the case. Clan exogamy may indeed be the expression 
of the feeling that marriage within the clan as such is incestuous; 
but it may also, as among the Toda and Blackfoot, be a secondary 
development, the fundamental fact being an objection to mar- 
riages between blood relatives. From Dr. Graebner's standpoint 
there is no reason to differentiate between the primary and the 
secondary type of clan exogamy. The form-criterion merely tells 
us that two groups are both exogamous; that in point of exogamy 
they are identical, and in so far may reasonably be supposed to 
be genetically related. So far as the criterion of quantity is con- 
cerned, nothing would be easier than to bolster up the parallel 
exogamy by other resemblances. Thus, the Crow social units, 
which exemplify the clan of "classical" ethnological literature in 
being exogamous in their own right, bear nicknames of similar 
type to that of the Blackfoot. Here again the identity of the 
facts compared is logical, while the facts we are really interested 
in studying are psychological. The exogamous conduct of the 


Theories and Theorists 

Blackfoot is inseparably linked with his feeling towards blood 
relatives; the exogamous conduct of the Crow is part of a quite 
distinct psychological complex. Only by disregarding the char- 
acteristic features of exogamy in these two instances do we get an 
identical Gedankending. 

In this connection it is interesting to discuss the two-phratry 
system (Zweiklassensystem) , as Graebner himself makes an ex- 
tensive use of this concept, suggesting, for instance, an historical 
connection between the two-pliratry organization of Oceania and 
that of the Northwest Coast Indians and the Iroquois. ^^ Before 
considering such a suggestion, we should have to be convinced 
that the term "two-phratry system" invariably labels the same 
phenomenon. Serious doubt is tlirown on such a supposition by a 
consideration of the data collected by Rivers among the Toda. 
In this tribe the numerical preponderance of one clan is such 
that its members can follow the exogamous rule only by marrying 
most of the members of the other clans, "leaving very few to in- 
termarry with one another." Out of 177 marriages, only 16 were 
between members of the other clans. As Rivers recognizes, there 
has thus developed the closest conceivable approximation to a 
two-phratry system. ^^ Yet this result has been achieved by unique 
historical causes quite distinct from those which brought about 
such a system where there are merely two intermarrying phratries 
without any lesser exogamous units. 

An instance of similar suggestiveness is furnished by the recent 
liistory of the Crow. A visitor to this tribe some forty years ago 
would have found the male members of the tribe grouped in two 
social units, — the Foxes and the Lumpwoods. Without any real 
feeling of mutual hostility, these two units were constantly pitted 
against each other; for example, taking opposite sides at games, 
and constantly attempting to outdo each other in warlike deeds. 
To a superficial observer this division would have appeared simi- 
lar to that of the Iroquois phratries, though, as a matter of fact, 
the Lumpwoods and Foxes were not social units with inherit- 
able membership, but military societies. At all events, even a 

" Graebner, "Die melanesische Bogenkultur . . . ," Anthropos, IV, 1021. 
" W. H. R. Rivers, "Totemism, an Analytical Study," Journal of American Folk- 
Lore, XXIII (1910), 246. 


Principle of Convergence in Ethnology 

more careful investigator might have been struck by the phe- 
nomenon as one comparable with the tendency to the formation 
of dual divisions, as evidenced in civilized life by the frequency of 
two dominant political parties. Nevertheless, forty years prior to 
the hypothetical investigator's advent, he would have found no 
less than eight societies of the same type.^^ A detailed study of 
the development of military societies among the Crow shows 
beyond a doubt that the presence of but two military organiza- 
tions forty years ago was not due to a primary dual organization, 
but came about solely through the elimination of the other or- 
ganizations. A comparison of the Crow conditions with those still 
more recently found among the Gros Ventre is of the utmost in- 
terest. In this tribe the old ceremonial grouping of the men in a 
rather large number of small companies representing probably 
six age-grades has been completely superseded by a division into 
two organizations, — the War Dancers and the Star Dancers. The 
tribal and social functions of these societies bear close resem- 
blance to those exercised by the Lumpwoods and Foxes of the 
Crow, and the spirit of rivalry is equally prominent in the Gros 
Ventre organizations. But while the dual grouping of the Crow 
men resulted from a process of elimination, precisely the reverse 
process took place among the Gros Ventre. The War Dance "is 
universally stated to be a recent importation from the Sioux, 
apparently within the present generation"; while the Star Dance 
is probably an old ceremony independent of the age-series. ^^ In 
the two cases under discussion, then, a dual grouping is beyond 
a doubt the result of convergent development. 

To revert to Graebner's own concepts, we may next consider 
his category of drums with skin drum-heads. -^^ He is careful to 
enumerate the several Oceanian forms; but as soon as his extra- 
Oceanian speculations begin, differences of form seem to become 
negligible. The skin drum of the West African culture-area is de- 

" Maximilian Prinz zu Wied, Reise in das innere Nord-America in den Jahren 
1832 bis 1834 (Coblenz, 1839), I, 401. 

^■^ A. L. Kroeber, Ethnology of the Gros Ventre, AMNH, Anthropological Papers, 
I (1908), 234-239. 

^' Graebner, "Ein Element von sehr typischer Verbreitung bieten zum Schlusse 
noch die Musikinstrumente in der einseitig bespannten, meist sanduhrformigen, 
bisweilen zylindrischen Felltrommel," Anthropos, IV, 770. 


Theories and Theorists 

scribed as one of the elements connecting it with Melanesian 
culture. It is said to appear with all the characteristic modes of 
securing the drum-head, — viz., by thongs, pegs, and wedges, 
— though the hourglass shape of the instrument is less frequent. ^'^ 

Probably it would be difficult to find a more offensive ex- 
ample of the misapplication of the form-criterion. The very ref- 
erence to the hourglass-shaped forms of Africa involves an er- 
ror of the worst kind. Graebner's authority defines the hourglass 
drum of Africa as composed of two skin-covered bowls connected 
by a cylindrical tube. Three sub-types are distinguished, of 
which two recall the shape of a dumbbell, while the third differs 
radically from the two others by the presence of four lugs and 
profuse decoration, and by the width of the connecting cylinder, 
which approximates that of the bowls. ^^ For convenience of de- 
scription, Ankermann is certainly justified in creating an hourglass 
type. But it would be unjustifiable to draw any inference as to 
genetic relations between the third and the two other sub-types; 
for quite apart from the elaborate decoration and the four lugs, 
the third sub-type is not at all similar to the dumbbell form. It 
is a psychological commonplace that even congruous geometrical 
forms may produce very different psychological effects. It is a 
fact known to field-workers in America that identical patterns are 
sometimes not recognized by the natives as identical if executed 
in different colors. A fortiori, we cannot assume without proof, 
that, where the divergence of form is very great, the native still 
assembles the varying forms under the same concept. Artifacts 
differ from organic forms in lacking an innate tendencv to vari- 
ability. If, therefore, we suppose that the lugged (Barotse-Am- 
boella) sub-type developed out of the dumbbell form, or vice 
versa, we introduce either the hypothesis that some external con- 
dition determined the change, or the psychological hypothesis that 
both forms were originally conceived as of one type. For neither 
of these suppositions is there the slightest foundation. 

If the foregoing argument applies within even a relatively 
continuous area, its force surely does not diminish when "hour- 
glass drums" of different continents are compared. Indeed, the 

"■'Ibid., pp. 1011 £ 

'* Bernhard Ankemiann, "Die afrikanischen Musikinstrumente," Ethnologisches 
Notizblatt, III (1901), 98 if., 53-55. 


Principle of Convergence in Ethnology 

hourglass drum of New Guinea, as described and pictured by 
Finsch, Biro, Sclilaginhaufen, and others, bears no resemblance 
to the African sub-types. We must regard the term "hourglass 
drum" as merely a convenient classificatory device by which may 
by described objects of diverse origin. The geometrical abstrac- 
tion defined by the term corresponds to no cultural reality; it 
develops in different areas by convergent evolution. 

As a matter of fact, the hourglass type which at least presents 
a semblance of morphological classification plays a very subor- 
dinate part in Graebner's treatment of the skin drum; for under 
the category of skin drums — and accordingly as evidence of a 
cultural connection between Oceania and North America — are 
cited the ordinary dancing-drum and the Midewiwin drum of the 
Ojibwa.^^ Thus the form-criterion is completely abandoned by 
its champion. 

It is true that Dr. Graebner, in his treatment of this subject, 
attaches considerable weight to the method of securing the drum- 
head, — whether by thongs, pegs, or wedges {Schnur-, Pfiock-, 
und Keilspannung) . This leads to an important question. How 
many ways of fastening a skin membrane to a drum are con- 
ceivable? Very little reflection is required to show that the num- 
ber is exceedingly limited. Indeed, the wedge system, being 
only a sub-type of the Schnur spannung, is not entitled to a special 
position on logical grounds, though from a comparative point of 
view it is incomparably the safest criterion of relationship. We 
must here apply what Dr. Goldenweiser has called, in conversa- 
tion with the author, "the principle of limited possibilities," which 
has recently been thus defined: "The theory of convergence 
claims that similar ways may (not must) be found. This would be 
a truism if there existed only one way of solving this problem; and 
convergence is obviously the more probable, the fewer the pos- 
sible solutions of the problem." ^^ In the case at hand, it cannot be 
taken as a sign of genetic connection that some African and some 
Oceanian tribes use pegs for fastening a drum-head, because 
the number of available ways is very small if classified in a man- 
ner that abstracts from all definite characteristics. 

This point is illustrated most clearly where the logical classifica- 

" Graebner, "Die melanesische Bogenkultur . . . ," p. 1021. 
^ Boas, op. cit., p. 807. 


Theories and Theorists 

tion involves a dichotomy of the universe. A well-known writer 
has discussed the origin myths of primitive folk, and found that 
some involve a theory of evolution, others one of special crea- 
tion. No sane ethnologist would infer from this that all the myths 
of either type were historically connected. To choose a somewhat 
more drastic illustration. Acquired biological traits must either 
be inherited or not inherited: consequently an expression of 
opinion, whether consciously or unconsciously bearing on the 
subject, must fall into either category. Many primitive tribes have 
myths recounting how in the remote past a certain animal met 
with some adventure which caused it to assume some biological 
peculiarity now noticeable in its descendants; nevertheless it 
would be absurd to accept this tacit assumption of transmission 
as a parallel of anti-Weismannism. Countless examples of a mode 
of classification rivaling in absurdity the hypothetical instance 
last cited are furnished by histories of philosophy. Too frequently 
the historian utterly neglects the processes by which conclusions 
are reached, and groups thinkers exclusively by the nature of 
their conclusions, which are labelled by descriptive catch- words. 
The identification of a philosopher as a monist or dualist, idealist 
or realist, is undoubtedly a labor-saving mode of characterization; 
but unfortunately it precludes a deeper comprehension of the 
thinker's philosophic individuality. A differentiation of social sys- 
tems on the basis of maternal and paternal descent, such as 
Graebner has undertaken, is justifiable within a limited area, 
where historical connections can be definitely demonstrated. Out- 
side such an area it can have no comparative significance, because 
descent cannot be reckoned otherwise than in either the maternal 
or the paternal line, or in both. 

The foregoing discussion has indicated the nature of the errors 
due to premature classification. The frequency of such errors, 
and the readiness with which they are committed, surely justify 
the greatest caution in identifying apparent homologies in the 
cultures of tribes not known to be historically related. The first 
question we must ask is, not how the trait could have travelled 
from one region to another, nor even whether it could have 

I 330 

Principle of Convergence in Ethnology 

originated independently through the psychic unity of mankind. 
Our first duty is rather to ascertain whether the resemblances 
are superficial or fundamental. For example, if we discover that 
the ipanang bali of the Sea Dyaks corresponds in the most strik- 
ing manner to the berdache of the Plains Indians,^^ we should not 
straightway identify the two institutions and invoke the principle 
of psychic unity or that of historical connection. Psychic unity 
would only explain the fact of a pathological variation, which 
seems to occur everywhere with a certain frequency. It does not 
explain why in but two particular areas this variation should lead 
to a marked social institution. Neither can historical connection 
be postulated in the absence of a tittle of evidence for either 
genetic relationship or transmission. The advocate of convergence 
in the sense here proposed will simply await a fuller determina- 
tion of the facts. If closer investigation should establish an abso- 
lute identity, the fact of identity would stand, but would stand 

But in many instances the identity of the cultural elements com- 
pared seems to be far more than an abstract possibility. The eye- 
ornament of the northwest coast of America is identical with 
that of Melanesia. For all practical purposes the star-shaped stone 
clubheads of New Guinea are identical with those from Peru. 
To put the case in the most general form, wherever we are deal- 
ing with objects which can be fully determined by an enumera- 
tion of their visible or sensible traits, there is the possibility of 
proving objective identity, as indicated by the examples just 
cited. However, there is an important consideration which can- 
not be neglected in this connection. The sensible traits of an 
ethnographic object may completely determine its character 
from the standpoint of the curiosity-dealer, but never from that 
of the scientific ethnologist.^^ For the latter a material object has 
a purely symbolical function: it represents a certain technique, 
an artistic style, a religious or social usage. In this sense it may 
be rightly said that "material" culture does not exist for the eth- 

^E. H. Gomes, Seventeen Years among the Sea Dyaks of Borneo (Philadelphia, 
1911), pp. 179 fF. 

"^ Cf. Franz Boas, "Some Principles of Museum Administration," Science, XXV 
(1907), 928. 


Theories and Theorists 

nologist, for the very word "culture" implies a psychological cor- 
relate, or rather determinant, of the material object. According 
to Pechuel-Loesche, the same representation of a human figure 
that in one West African specimen is nothing but a product of 
art industry becomes, when endowed with certain magical 
powers by virtue of incantations or the application of sacred sub- 
stances, a fetich. Exactly the same purpose, however, may be 
served in the same tribes by the most inconspicuous objects of 
nature. A purely objective comparison would here lead to an 
utterly erroneous classification. It would wrest the factors studied 
out of their organic context in quite the same way as an identifica- 
tion of the cultural traits discussed in the preceding section; it 
would neglect the very factors that we are most interested in 

As has been pointed out by American archaeologists, the appli- 
cation of the form-criterion is insufBcient in determining the antiq- 
uity of an archaeological object; for the latter may not be at all the 
completed object designed by the worker, but a mere "reject." ~^ 
Yet objectively the rejects coincide absolutely with the finished 
products of a lower culture. The difi^erence lies in the cultural 
contexts of which the objects are elements: the resemblance may 
be perfect from a purely external standpoint; nevertheless it repre- 
sents, in Ehrenreich's terminology, not a genuine convergence, 
but a false analogy. A most suggestive fact pointing in the same 
direction has been ascertained in Central Austrafia. The natives 
of this area use implements, some of which fall morphologically 
under the category of paleoliths, while others are neoliths. In- 
vestigation has shown that this morphological difference is a 
direct result of the material available for manufacture. Where 
diorite is available, the natives manufacture "neolithic" ground 
axes, in other cases they make flaked implements practically as 
crude as those of the ancient Tasmanians.^^ The manufacture of 
"neolithic" implements in Central Australia and elsewhere thus 
forms another instance of convergence, — a classificatory resem- 
blance due to heterogeneous conditions. It is true that Graebner 

^ J. Alden Mason, The Origins of Invention, p. 124. 

^ Baldwin Spencer and F. J. Gillen, The Northern Tribes of Central Australia 
( London and New York, 1904 ) , p. 635. 


Principle of Convergence in Ethnology 

does not ignore the possible influence of material on form,^^ but 
he fails to show under what circumstances the ethnologist should 
seek to correlate morphological resemblance with the nature 
of the material. The form-criterion by itself does not tell us that 
diorite lends itself to "neolithic" workmanship, that bamboo bows 
are necessarily flat, that basalt furnishes the only material avail- 
able for axe-manufacture in certain regions. Under what con- 
ditions should we be satisfied with formal coincidence as a proof 
of genetic relationship, and under what conditions should we 
inquire as to the possible influence of the available material? 

The case of the eye-ornament adds force to the general argu- 
ment. As Graebner might have learned from Ehrenreich's article 
(op. cit., p. 179), Boas has shown that the eye-ornament of North- 
western America results from a peculiar style of art, which, so far 
as we know, does not occur in Oceania; that is to say, the objec- 
tive identity is again deceptive, because it is an identity estab- 
lished by wresting a part of the phenomenon studied ( the visible 
pattern ) from the midst of its cultural context. Here it must again 
be stated that Graebner does not unqualifiedly uphold the om- 
nipotence of the morphological principle. He rejects Von Lusch- 
an's speculations on the head-rests of New Guinea; he regards 
Schurtz's theories of the eye-ornament as "weniger phantastisch, 
aber doch auch iibers Ziel geschossen"; he stigmatizes Stucken's 
attempt to trace all celestial myths to Babylon as an example of 
the neglect or unmethodical application of the form-criterion 
(p. 118). Unfortunately, he does not explain what is meant by 
an unmethodical or fantastic application of the form-criterion. 
As has been shown, the criterion of quantity is a measure of the 
historical connection between cultures, but can never decide as 
to the identity of doubtful traits. If all the other elements of 
Oceanian and northwest American culture were identical, the fact 
would prove nothing as to the identity of the eye-ornament in 
the two areas. 

We are not always, indeed we are very rarely, in the fortunate 
position of knowing most of the determining conditions of an 
ethnological phenomenon. In the case of the rejects, of the cen- 
tral Australian "neoliths," and of the eye-ornament, we happen 

^ Graebner, Methode der Ethnologie, pp. 145, 117. 


Theories and Theorists 

to be in possession of the facts; and from these instances we learn 
that morphological identity may give presumptive, but does not 
give conclusive, evidence of genetic relationship. It is conceivable 
that if we could determine the history of the South American 
paddles, which Graebner connects with Indonesian and Mela- 
nesian patterns,^*^ we should find them to be genetically related; 
but we cannot bar the other logical possibility of independent 
origin, for it is hkewise conceivable that each of the homologous 
features of the paddies originated from distinct motives and dis- 
tinct conditions. 

The doctrine of convergence, as here advocated, is not dogmatic, 
but methodological and critical. It does not deny that simple 
ethnological phenomena may arise independently in different 
regions of the globe, nor does it deny that diffusion of cultural 
elements has played an important part. It does not even repudiate 
the abstract possibility of the independent origin of complex 
phenomena (genuine convergence of Ehrenreich), though so far 
the demonstration of identities of such a character seems insuf- 
ficient, and their existence would be unintelligible. The view here 
propounded demands simply that where the principle of psychic 
unity cannot be applied, and where paths of diffusion cannot be 
definitely indicated, we must first inquire whether the supposed 
identities are really such, or become such only by abstracting 
from the psychological context in wliich they occur, and which 
determines them, — whether, that is to say, we are comparing 
cultural realities, or merely figments of our logical modes of clas- 
sification. A rapid survey of the field has sufficed to show that 
in many cases where some would invoke the principle of psychic 
unity, and others that of historic connection, the problem is an 
apparent one, which vanishes with a better knowledge and classifi- 
cation of the facts. 

Dr. Graebner's ambitious attempt to trace historical connections 
between remote areas cannot be dismissed wholesale, on the basis 
of the foregoing criticisms. What has been shown is simply the 

^'' Ibid., p. 145; idem, "Die Melanesische Bogenkultur . . . ," Anthropos, IV, 
763, 1016, 1021. 


Principle of Convergence in Ethnology 

necessity for a critical use of ethnological concepts, and their oc- 
casionally quite uncritical use of Graebner. Even tangible speci- 
mens, it appears, cannot be studied apart from the culture of 
which they are a product. In the investigation of social and re- 
ligious usages, where the subject-matter is itself psychological, 
the exclusive consideration of the form-criterion, to the detriment 
of the subjective factors involved, can lead only to disastrous re- 
sults. Ethnology is a relatively young science, and it is natural 
that the mode of classification in vogue among ethnologists should 
have a pre-scientific tang. But the time has come to recognize 
that an ethnologist who identifies a two-class system in Australia 
with a two-class system in America, or totemism among the North- 
western Indians with totemism in Melanesia, sinks to the level of 
a zoologist who should class whales with fishes, and bats with 



Ceremonialism in North America 

In delimiting the range of cultural phenomena to 
which this paper will be confined, it is impossible to adhere to any 
of the current definitions of "ceremony" or "ceremonial." A set mode 
of procedure is characteristic of every phase of primitive behavior, 
and thus it is justifiable to speak of birth, puberty, death, war 
ceremonies, etc. An article on "ceremonialism" in this sense would 
needs center in a discussion of the psychology of routine. When, 
however, Americanists speak of "ceremoniahsm," they generally 
associate with the term a more or less definite content of stereo- 
typed form. Performances such as the Snake Dance of Pueblo 
peoples, the Sun Dance of the Plains, the Midewiwin of the 
Woodland area, are examples par excellence of what is commonly 
understood by a "ceremony." These performances are not in- 
dividual, but collective undertakings; and, even where they 
hardly fall under the category of "religious observances" or "sol- 
emn rites," they are uniformly more than mere attempts at social 
amusement. As Indian dances are often performed for a serious 
purpose, or at least form elements of complexes of a serious 
character, the terms "dance" and "ceremony" are sometimes used 
interchangeably. This loose usage is as undesirable as the fre- 

American Anthropologist, XVI (October-December, 1914), 602-631. 


Ceremonialism in North America 

quent identification of the problem of ceremonialism with that 
of organizations. There are North American dances performed 
exclusively as a matter of amusement, and there are organizations 
corresponding to our clubs rather than to ceremonial bodies. Ele- 
ments of similarity may necessitate joint consideration of the 
ceremonial and non-ceremonial dances and societies; but it may 
be well to state that, in dealing with "ceremonialism," we start 
primarily from a consideration of solemn collective performances 
with an avowedly serious purpose, and shall include only such 
other phenomena as are historically or psychologically related to 
"ceremonialism" as thus defined. 

Having regard to the limitation of space, a descriptive account 
of ceremonial activity in North America is out of the question 
here. I shall therefore merely enumerate the most important cere- 
monies in the several culture provinces, and shall then select for 
discussion a number of problems that arise from the considera- 
tion of our ceremonial data. 

In the Eastern Woodland area, the Midewiwin looms as the 
most important ceremony of the Algonquian tribes, though its 
sphere of influence extended to several Siouan peoples, includ- 
ing some inhabiting the Plains. It was the property of a secret 
society, membership in which was preceded by a formal initiation. 
A shooting performance, either by way of initiating the novice or 
merely as a shamanistic practice, forms the most obvious objec- 
tive bond between the forms of the ceremony as practised by 
the several tribes; while the interpretation of the aim of the cere- 
mony varies.^ The Iroquois also had a number of secret ceremo- 
nial organizations of as yet little understood character, of which 
may be mentioned the Little Water Fraternity and the False Face 
Society; the performances of the latter being characterized by 
the use of grotesquely carved face-masks. In addition, there was 
a series of tribal seasonal festivals, ostensibly in the nature of 
thanksgiving celebrations, held annually at such periods as the 
first flowing of the maple-sap, the planting and the ripening of the 

^William Jones, "Central Algonkian," Annual Archaeological Report for 1905 
(Toronto, 1906), p. 146; Paul Radin, "The Ritual and Significance of the Win- 
nebago Medicine Dance," Journal of American Folk-Lore, XXIV (1911), 149- 
208; W. J. Hoffman, "The Midewiwin or 'Grand Medicine Society' of the Ojibwa," 
BAE, 7th Ann. Kept. ( 1889), pp. 149-500. 


Theories and Theorists 

com, etc. These ceremonies, as well as the seven-days' New 
Year's Jubilee, correspond in a way to the spectacular composite 
perfomiances of other areas in wliich religious practices are 
combined with entertainments of various forms .^ 

In the Southeast all other dances were completely overshad- 
owed by the annual several-days' (from four to eight) festival 
known as the "Busk," and celebrated on the first ripening of the 
crops. The public making of new fire, the scarification of the men, 
and the taking of an emetic, are among the noteworthy objective 
features. The new-fire ceremony, as pointed out by Speck, has 
analogies not only in the Southwest, but even in Mexico; and the 
taking of an emetic is shared with some southern Plains tribes 
and the Pueblo Indians.^ 

In the Plains area, ceremonial activity attained a very high de- 
gree of development, though this was shared in very unequal 
measure by the several tribes. The Sun Dance, the great tribal 
performance of most of the inhabitants of the area, will be dis- 
cussed below. Other ceremonial perfonnances of wide distribu- 
tion center in the rites connected with sacred bundles of restricted 
ownership. The widely diffused medicine-pipe ceremonials, the 
sacred-bundle rites of the Blackfeet, and the shrine performances 
of the Hidatsa, may serve as examples. There are mimetic ani- 
mal dances, those in imitation of the buffalo occurring in vary- 
ing guise and with varying raison d'etre, such as the luring of 
the game. Some of the last-mentioned category of perfonnances 
are the property of individuals who have experienced a vision of 
the same supernatural animal. Military and age societies, though 
in certain tribes wholly or predominantly secular, assume in others 
a markedly ceremonial aspect.'^ 

^ A. C. Parker and H. M. Converse, Myths and Legends of the Neio York State 
Iroquois, New York State Museum, Bull. 125 ( 1908), pp. 74 ff., 149 ff.; L. H. Mor- 
gan, League of the Hodenosaunee, or Iroquois (Rochester, 1854), pp. 187-222, 

^ F. G. Speck, Ethnology of the Yuchi Indians, University of Pennsylvania, Anthro- 
pological Publications, University' Museum, I (1909), 112—131. 

* G. A. Dorsey and A. L. Kroeber, The Arapaho Sun Dance, Field Columbian 
Museum, Anthropological Series, V (1903), 1-228; G. A. Dorsey, The Cheyenne, 
same ser., IX ( 1905), 143-358; J. O. Dorsey, "A Study of Siouan Cults," BAE, 11th 
Ann. Rept. (1894), pp. 351-544; AUce C. Fletcher, "The Hako: A Pawnee Cere- 
mony," BAE, 22d Ann. Rept. ( 1904), Part 2; A. C. Fletcher and Francis La Flesche, 


Ceremonialism in NoHh America 

Among the Southwestern Indians, North American ceremonial- 
ism attains its high-water mark. There is a profusion of rituaUs- 
tic externals, — wooden or sand-painted altars, prayer-offerings, 
masks, sacred e£Bgies, and the like, — and esoteric fraternities 
perform elaborate ceremonies in order to heal the sick, or for 
the ostensible purpose of promoting the public welfare by effect- 
ing adequate rainfall or insuring success in the chase or war. 
These performances resemble the Iroquois festivals and the Plains 
Indian Sun Dance in being composite phenomena in which 
strictly religious features are blended with games, clownish pro- 
cedure, and what not. The Hopi and Zuiii ceremonies further 
recall the Iroquois festivals in being calendric; that is, fol- 
lowing one another in fixed sequence at stated seasons of the 

On the Northwest coast and its immediate hinterland we find 
the potlatch festival, involving a generous distribution of prop- 
erty by the host that entails a return distribution of gifts at a 
high rate of interest. Upon this secular basis there have been 
engrafted, among the northern tribes of the area, ceremonial con- 
cepts derived from the Winter Ritual of the northern Kwakiutl, 
from whose territory they have likewise extended southward. 
The Winter Ritual is founded on the novice's acquisition of a su- 
pernatural protector, whose character is in a measure predeter- 
mined by his family affiliations, or rather restricted by his family's 
supernatural property rights. During the winter, community of 
guardian spirits forms the bond of association, superseding family 
ties, and creating temporarily a number of ritualistic societies. 

"The Omaha Tribe," BAE, 27th Ann. Rept. (1911); A. L. Kroeber, The Arapaho, 
AMNH, Bull, XVIII (1902-1907), 1-229, 279-454; idem. Ethnology of the Gros 
Ventre, AMNH, Anthropological Papers, I (1908) 141-282; R. H. Lowie, The 
Assiniboine, same ser., IV (1909), 1-270; idem. Societies of the Crow, Hidatsa and 
Mandan Indians, same ser., XI ( 1913), 143-358; Clark Wissler, Ceremonial Bundles 
of the Blackfoot Indians, same ser., VII (1912), 65-289; idem. Societies and Dance 
Associations of the Blackfoot Indians, same ser., XI ( 1913), 363-460. 

^ J. W. Fewkes, "The Snake Ceremonials at Walpi," Journal of American Eth- 
nology and Archaeology, IV (1894); idem, "The Group of Tusayan Ceremonials 
Called Katcinas," BAE, 15th Ann. Rept. (1897), pp. 251-313; Washington Mat- 
thews, The Night Chant, a Navaho Ceremony, AMNH, Memoirs, VI ( 1902); Matilda 
C. Stevenson, "The Sia," BAE, 11th Ann. Rept. (1894), pp. 16, 69-131; idem, 
"The Zuni Indians," BAE, 23d Ann. Rept. (1904), pp. 62-283. 


Theories and Theorists 

The ritual purports to portray the novice's abduction by the 
guardian spirits, their return to the village, and their restoration 
to a normal condition. In reality it is a compound of these ele- 
ments with potlatch incidents, sleight-of-hand exhibitions, clown- 
ish activity, and so forth.® 

Among the Eskimo unaffected by neighboring Indian peoples, 
ceremonialism apart from shamanistic practices is but slightly 
developed. The Central Eskimo have an annual festival that pur- 
ports to effect the home-sending of the deity protecting the sea- 
mammals, and during which the shaman purges this deity's body 
by removing the effects of transgressed taboos. The appearance 
of masked performers impersonating the divinity and other spirits 
is a noteworthy trait of this ceremony.'^ 

Paucity of ceremonial is a trait shared by the inhabitants of 
the Mackenzie area, the Plateau region, and California, all of 
whom present the least highly developed form of North American 
culture. Professor Kroeber has pointed out that the simpler the 
stage of culture the more important is the shaman.^ The statement 
might be extended from shamanistic practices to those practically 
universal observances connected with such events as birth, pu- 
berty, individual acquisition of supernatural power, and death. 
They, like the shamanistic functions in Kroeber's characteriza- 
tion, tend to become, "relatively to the total mass of thought and 
action of a people, less and less important." It thus seems possible 
to consider ceremonialism par excellence, as defined above and 
treated by preference in this article, a relatively recent trait super- 
imposed on a series of simple routine procedures of the type 
just mentioned. The culture of the Mackenzie River people is 
relatively little known, but the prominence of shamanism and 

" Franz Boas, "The Social Organization and the Secret Societies of the Kwakiutl 
Indians," USNM, Report for 1895, pp. 311-737; idem. The Mijthology of the Bella 
Coola Indians, Pubhcations of the Jesup North Pacific Expedition, Vol. I ( 1898 ) ; 
J. R. Swanton, Contribution to the Ethnology of the Haida, same ser., Vol. V 
(1905); idem, "Social Conditions, Beliefs and Linguistic Relationship of the Tlingit 
Indians," BAE, 26th Ann. Kept. (1908), pp. 391-485. 

' Franz Boas, "The Central Eskimo," BAE, 6th Ann. Rept. ( 1888), pp. 583-609; 
idem. The Eskimo of Baffin Land and Hudson Bay, AMNH Bull., XV ( 1907), 119 ff., 
489 IF. 

* A. L. Kroeber, The Religion of the Indians of California, Univ. Calif, Publ. 
Amer. Arch, and Ethn., IV ( 1907), 327. 


Ceremonialism in North America 

sleight-of-hand tricks appears clearly from Hearne's and Petitot's 
accounts; ^ and among the Thompson River Indians the puberty 
ceremonials loom as a very important cultural feature, ^*^ Sha- 
manism with its correlated practices, and puberty rites, are known 
in other areas, but they are often eclipsed by the doings of eso- 
teric brotherhoods and other spectacular performances. This 
is merely grazing a significant problem; and it must be clearly 
understood that, even in the ruder North American cultures, 
phenomena comparable to the more impressive ceremonials of 
other regions are not wholly lacking. Thus the Ute and related 
Shoshoneans celebrate an annual spring festival known as the 
"Bear Dance"; ^^ a series of winter dances with ceremonial rai- 
ment occurs among the Central Californian Maidu; and other 
Californian tribes have public annual mourning ceremonies and 
the semblance of a secret society formed by initiated male tribes- 
men.^^ The occurrence of these elements even in the simplest 
cultures seems to indicate rather clearly that the differences in 
ceremonial development are not correlated with psychological 
differences, but rather with differences in the manner of combin- 
ing and multiplying elements of general distribution. A hint as 
to the luxurious growth of ceremonialism in certain areas will be 
found in the section on "Ceremonial Patterns," though why a cer- 
tain feature extant in a number of regions should become a pat- 
tern in one tribe, and fail to become one in others, remains ob- 

Another question, which it is impossible more than to hint at 
here, relates to the distribution of ceremonial traits less widely 
diffused than those just dealt with. Thus ceremonial public con- 
fession is a trait shared by the Eskimo ^^ with the Iroquois ^^ and 
the northern Athapascans.^^ In this case geographical considera- 

° Samuel Heame, A Journey from Prince of Wales' Fort in Hudson's Bay to the 
Northern Ocean (London, 1795), pp. 191-194; E. Petitot, Traditions indienncs 
du Canada Nord-Ouest (Paris, 1886), pp. 434-436. 

^° James Teit, The Thompson Indians of British Columbia, Publications of tlie 
Jesup North Pacific Expedition, V (1905), 311-321. 

" My own field information. 

" Kroeber, The Religion of the Indians of California, pp. 334 IF. 

" Boas, The Eskimo of Baffin Land and Hudson Bay, p. 121. 

" L. H. Morgan, op. cit., p. 187. 

'' Petitot, op. cit., p. 435. 


Theories and Theorists 

tions point with overwhelming force to an explanation by his- 
torical contact. The above-mentioned instance of the new-fire 
ceremony forms perhaps an almost equally good case in point; 
but in other cases the matter is less certain, though odd features 
of capricious distribution haunt the mind with visions of possible 
historical connection. Thus Boas refers to the rather striking 
analogies between the tortures of the Kwakiutl War Dance and 
the Plains Indian Sun Dance. ^*^ The phenomenon of ceremonial 
buflFoonery that crops up among the Iroquois, the western Ojibwa, 
many of the Plains tribes and Pueblo Indians, as well as in Cah- 
fornia and on the Northwest coast, presents probably too general 
a similarity (except among tribes obviously in contact with one 
another) to be considered of historical significance. Nevertheless 
some specific analogies are puzzling. Thus the Tlingit have so 
distinctive an element of Plains Indian clownishness as the use of 
"backward speech"; that is, expression of the exact opposite of the 
intended meaning. -^^ Only a much fuller knowledge of the dis- 
tribution of ceremonial elements and complexes will help us esti- 
mate the relative value of the theories of historical contact and 
independent development in such concrete instances. For the 
time being, it will be well to regard historical contact as estab- 
lished only in the clearest cases, though these are by no means 
few (see below, "Diffusion of Ceremonials"). 

In many cases a ceremony is derived by the natives from a myth 
accounting for its origin. Native statements, however interesting 
in themselves, cannot of course be taken as objective historical 
fact. Hence arises the question, Is the myth the primary phe- 
nomenon on which the ceremony is founded, or is it merely a sec- 
ondary explanation of the origin of a preexisting ceremony? A 
considerable amount of information bearing on this problem has 
been recorded; here only enough can be presented to illustrate 
essential principles. 

The Crows and Blackfeet share a ceremonial planting of Sacred 
Tobacco. As this performance has not been found among other 

' Boas, "The Social Organization ... of the Kwakiutl Indians," pp. 495, 661. 
Swanton, "Social Condition ... of the Tlingit Indians," p. 440. 


Ceremonialism in North America 

tribes of this area, and as there are similarities of detail, the single 
origin of the common features of the ceremonies as performed by 
the two tribes is certain. Among the Blackfeet, however, the 
Sacred Tobacco forms part and parcel of the Beaver Medicine 
Bundle. This is in its entirety derived from a Beaver, who, after 
luring away a Blackfoot's wife, indemnified the husband by send- 
ing the woman back with the Beaver Bundle. ^^ The Crows, on 
the other hand, do not associate their Tobacco with the beaver, 
but identify it with the stars. According to the most popular ver- 
sion, the discovery of the Tobacco dates back to the period of their 
legendary separation from the Hidatsa, when one of two brothers 
was adopted by the stars, blessed with the vision of the Tobacco, 
and instructed as to the ceremonial planting. The same ritualistic 
features are thus associated with two distinct myths in the two 
tribes; hence at least one of the myths is certainly secondary, 
which establishes in principle the possibility of such a secondary 
association. For the secret ceremonials of the Northwest coast 
of North America, a corresponding conclusion was long ago 
drawn by Professor Boas. Of the several tribes sharing the cere- 
monies in question, some derive their performances from the 
wolves, others from heaven, still others from the cannibal spirit 
or from a bear. In all cases but one, the explanation must be 
secondary, and, with the possibility of such explanation estab- 
lished, it becomes psychologically justifiable to treat tlie residual 
case as falling under the same category: the ritualistic myth is an 
aetiological myth. Ehrenreich has duly emphasized the oc- 
currence of demonstrably secondary connection between ritual 
and myth in North America; and, since the rituals and myths of 
this continent are better known than those of any other area of 
equal magnitude, he rightly insists that the conclusions derived 
from this basis have general significance for the problem of the 
relationship of these associated elements. ^^ 

Boas and Ehrenreich not only strengthen the case for secondary 
connection, but also demonstrate the workings of the aetiological 

" Clark Wissler, Mythologtj of the Blackfoot Indians, AMNH, Anthropological 
Papers, II (1908), 1-164. 

" Paul Ehrenreich, Die allgemeine Mt/thologie und ihre ethnologishchen Grund- 
lagen (Leipzig, 1910), p. 84. 


Theories and Theorists 

instinct by proving that in not a few cases a ritual is accounted for 
in a single tribe by attaching it to a folk-tale or folk-tale episode 
of very wide distribution. In such instances the question of the 
priority of the tale or ritual is, of course, immaterial: there is 
secondary association of previously independent units. 

Thus, among the Heiltsuk alone, the story of a woman who gave 
birth to dogs is used to explain the establishment of the Cannibal 
Society. As this tale is found without any ceremonial associations 
among the Eskimo, all the northern Athapascans, and all the 
Northwest coast Indians, its secondary application to the Heiltsuk 
ritual is manifest. In other words, not only is the same ritual ex- 
plained by dijfferent myths in diflPerent tribes, but, in the attempt 
to account for the origin of the ritual, there is a tendency to use 
popular tales that come to hand.^° This tendency, it may be noted, 
is strongly developed in other regions of the continent. The 
Hidatsa and Mandan associated the custom of planting certain 
offerings by the bank of the Missouri with the tale of the young 
man who ate of the flesh of a snake, became transformed into a 
snake, and was carried to the Missouri by his comrade.^^ Accord- 
ing to my own field data, these off^erings formed part of the Hidatsa 
Missouri River ceremony, one of the sacred rituals of the tribe. 
Similarly, the Bird ceremonial of the same tribe is connected 
with the exceedingly widespread story of the thunderbird's an- 
tagonism to a water-monster. Examples of this type certainly 
seem to justify in considerable measure Ehrenreich's conclusion: 
"Jedenfalls liegen der Kegel nach einem Kultmythus schon ander- 
weitig bekannte Stoffe oder in anderen Verbindungen vorkom- 
mende mythische Elemente zugrunde. Was das Ritual dem hin- 
zufiigt, ist ausseres Beiwerk, als Anpassung zu bestimmtem 

There are many instances, however, where the connection be- 
tween ritual and myth is of a more intimate nature. The Black- 
foot myth of the Beaver Bundle, quoted above, which forms the 

^ Boas, "Social Organization ... of the Kwakiutl Indians," pp. 662-664; idem, 
The Mythology of the Bella Coola Indians, p. 126. 

^ Maximilian, Prinz von Wied-Newied, Reise in das innere Nord-America 
(Coblenz, 1839), II, 184-186, 230-234. The tale, without ritualistic associations, 
occurs among the Assiniboine, Arapaho, Gros Ventre, Crow, Omaha, and Arikara. 
See Lowie, The Assiniboine, p. 181. 


Ceremonialism in North Annerica 

pattern for a series of other ritualistic myths, may serve as an 
example. "In most ceremonies," writes Wissler, "the origin of the 
ritual is regarded as the result of a personal relation between its 
first owner and its supernatural giver; each ceremony or demon- 
stration of the ritual being a reproduction of this formal trans- 
fer." ^^ This notion is so strongly developed among the Hidatsa 
that, whenever one of my informants was unable to recount the 
vision through which knowledge of a particular ceremony was 
derived, he at once suggested that the ceremony must be of 
foreign origin. Substantially there is no difference between the 
origin myths and the accounts by men still living of such visions 
as explain the institution of recent ceremonies: both recount the 
meeting with the visitant, his ceremonial gifts, and relevant in- 
structions. The only difference lies in the fact that stories of the 
first class have already, while those of the second class have not 
yet, become part of the traditional lore of the tribe, or clan, or 
society. Again, the secondary character of the myth is at once 
manifest: no tribe could develop a story explaining ceremonial 
details ( any more than an individual could have a vision of such 
ritualistic proceedings), unless such ceremonial features already 
formed part of the tribal consciousness. The myth simply recites 
the preexisting ritual, and projects it into the past. 

There is, of course, nothing in the nature of human psychology 
that would prevent myths from being dramatized in ceremony. 
It is simply an empirical fact that in North America such drama- 
tization, if not wholly absent, is certainly subordinate in impor- 
tance to the aetiological utilization of the myth. The Midewiwin 
ceremony does not dramatize the doings of Manabush and his 
brother; but the celebrants recite the story and add to it an ac- 
count of the origin of their own doings. The Omaha Shell Society 
interpret the ceremonial shooting practised by members as a 
dramatic representation of the shooting of four children in the 
Origin Myth; but, as Radin has shown,^^ the shooting ceremony 
is so widespread a feature in other tribes that it cannot have 
originated from this particular tale. The Okipa performers do not 
enact their tale of a flood, but use that tale as a partial explanation 

^ Wissler, Mythology of the Blackfoot Indians, p. 13. 
^ Paul Radin, op. cit., p. 182. 


Theories and Theorists 

of their annual festival. A secondary reflex effect of the myth on 
the ritual and its symbolism is of course undeniable. Thus in the 
Okipa we do find an actor impersonating the mythic hero, Nii'mak- 
maxana; but, while the actor narrates the tale of the flood, he does 
not, so far as we can judge, perform the actions of his prototype at 
the time of the flood or on any other occasion. Similarly, among 
the Hidatsa, the hero-trickster figures in many ceremonial per- 
formances; but he does not act out his heroic or clownish ex- 
ploits.^* Again, among the Bellacoola, the kusiut ceremonial 
appears to the native mind as a dramatic representation of legend- 
ary happenings. As a matter of fact, we do meet with impersona- 
tions of the deities of the Bellacoola pantheon; but the essential 
elements of the ceremonial, such as the cannibalistic practices, 
have an origin, not in the highly specialized Bellacoola mythology, 
but in actual observances shared in recent times by a number 
of Northwest coast tribes, and connected in part with war customs. 
So among the Hopi the episodes of the legends associated with 
ceremonials do not determine at all definitely the sequence of 
ceremonial procedure; here also the ritual appears as a less variable 
and as a preexisting feature.^^ Finally may be mentioned the 
Mohave case. Here the ceremonies not connected with mourning 
"consist essentially of long series of songs, occupying one or more 
nights in the recital, which recount, in part directly but more 
often by allusion, an important myth. At times the myth is ac- 
tually related in the intervals between the songs. In some cases, 
dancing by men or women accompanies the singing; but this is 
never spectacular, and in many cases is entirely lacking." ^^ But, 
though the prominence of the myth is here so great that the 
ceremonies in question are only ceremonial recitations of myths, 
this very fact obviously precludes dramatization of the mythic 

In the Plains area, the diffusion of ceremonies is in some cases 
not merely a plausible hypothesis, but an historical fact. No one 

^ G. H. Pepper and G. L. Wilson, "An Hidatsa Shrine and the Behefs Respecting 
It," AMNH, Memoirs, II (1908), 320. 

^ Fewkes, "The Group . . . Called Katcinas," pp. 253 ff. 
^' Kroeber, The Religion of the Indians of California, p. 340. 


Ceremonialism in North America 

could doubt that the Hot Dance of the Arikara, Ruptare Mandan, 
and Hidatsa (involving in each instance the plunging of the per- 
formers' arms into scalding hot water) must have been derived 
from a common source. But we have in addition Maximilian's as- 
surance that the ceremony was obtained by the Hidatsa from the 
Arikara. ^^ Lewis and Clark (1804) mention ceremonial fool- 
hardiness as a feature borrowed by the Dakota from the Crows. ^^ 
Within the memory of middle-aged men at least, two ceremonies 
have been introduced into the northern Plains from the south. 
The peyote cult, which is found among the Tepehuane, Huichol, 
and Tarahumare of Mexico, flourishes among the Kiowa and Co- 
manche, and has thence traveled northward to the Arapaho, 
and even to the Winnebago."'' The Grass Dance was introduced 
among the Crows by the Hidatsa about 1878; among the Black- 
feet by the Gros Ventre, about 1883; among the Flathead by 
the Piegan, in quite recent times, ^"^ It seems to have originated 
among the Omaha and cognate tribes, including the Ponca, Osage, 
Iowa, and Oto.^^ In addition to the tribes already mentioned, its 
occurrence has been noted among the Pawnee, Dakota, and As- 
siniboin. Other unexceptionable instances are numerous. Thus a 
Medicine Pipe Dance of the Pawnee hako type was adopted by 
the Crows from the Hidatsa during the second half of the nine- 
teenth century; and the Hidatsa remember that their Medicine 
Pipe ceremony was in turn derived from the Arikara. A sacred 
Horse Dance practised by the River Crows was secured from the 
Assiniboin. The same division of the Crows adopted a Crazy 
Dog Society from the Hidatsa about thirty-five years ago. To pass 
to another area, the Kwakiutl proper ascribe the origin of their 
cannibalistic ceremonial to the Heiltsuk, from whom they derived 
the practice in approximately 1835; while the Tsimshian derive 

^^ Maximilian, op. cit., II, 144. 

^ Lewis and Clark, Original Journals of the Lewis and Clark Expedition ( Thwaites 
ed.; New York, 1904-1905), p. 340. 

^^ Kroeber, The Arapaho, p. 320; idem. Handbook of the h^dians of California, 
BAE, Bull. 78 (1925); Paul Radin, "A Sketch of the Peyote Cult of the Winne- 
bago," Journal of Religious Psychology, III (1914), 1-22. 

'"' Lowie, Societies of the Crow, Hidatsa and Mandan Indians, p. 200; Wissler, 
Societies and Dance Associations of the Blackfoot Indians, p. 451. 

"1 A. C. Fletcher and F. La Flesche, "The Omalia Tribe," BAE, 27th Ann. Kept. 
(1911), p. 459. 


Theories and Theorists 

a corresponding custom from the same som^ce, whence it reached 
them probably ten years before.^^ While native tradition is often 
untrustworthy, the date set by it in these instances is so recent 
that scepticism is hardly in place. This is especially true, since 
linguistic evidence supports the account of the Indians; for practi- 
cally all the names applied to the Tsimshian performances are 
derived from the Kwakiutl, and the characteristic cry of the can- 
nibal is likewise a Kwakiutl word.^^ 

The foregoing instances, which could be considerably mul- 
tiplied, illustrate diffusion as an observed or recollected historical 
phenomenon. Even in the absence of such direct evidence, how- 
ever, the theory of diflFusion is in many cases inevitable. Among 
the graded ceremonies of the Gros Ventre, the lowest is a Fly 
Dance, which is said to have been instituted by a Mosquito; the 
members imitated mosquitoes, pursuing people and pricking 
them with spines and claws. The lowest of the graded Blackfeet 
ceremonies recorded by Maximilian in the early thirties of the 
nineteenth century was likewise practised by a Mosquito So- 
ciety, whose members imitated mosquitoes, maltreating their 
fellow-tribesmen with eagle-claw wristlets. ^^ The coincidence is 
so complete in this instance, that a common origin is certain, es- 
pecially since the Blackfeet and Gros Ventre have been in intimate 
contact with each other, and since the only other people known to 
have had a Mosquito ceremony, the Sarsi, have also been closely 
associated with the Blackfeet. In the case at hand, we are even 
able to go a step farther, and ascertain not merely the fact, 
but the direction, of the diffusion process. The Gros Ventre are 
linguistically most closely allied with the Arapaho, with whom 
they once lived, and whose ceremonial system presents striking 
resemblances to their own. The presence of a Mosquito Dance 
among the Gros Ventre constitutes one of the glaring disparities 
amidst otherwise far-reaching likenesses: we may therefore reason- 
ably infer that the difference resulted from the adoption of the 
Blackfeet Mosquito Dance by the Gros Ventre subsequent to 
their separation from the Arapaho. 

Boas, "The Social Organization ... of the Kwakiutl Indians," p. 664. 

Ibid., p. 652. 

Lowie, The Assiniboine, p. 82. 


Ceremonialism in NoHh America 

In other cases we must be content to infer the mere fact of 
diffusion from the observed homologies. For example, the Arapaho 
and Cheyenne have each a Dog organization with four scarf -wear- 
ing oflScers pledged to bravery, and characterized by the same 
ceremonial regalia, such as dew-claw rattles, feather head-dresses, 
and eagle-bone whistles. The union of these logically quite un- 
related features in adjoining tribes establishes beyond doubt a 
common origin; but I am not acquainted with any specific data 
that would indicate whether the Arapaho borrowed from the 
Cheyenne, or vice versa. Cases of this type are exceedingly com- 
mon in every one of the principal culture areas; and where simi- 
larities extend beyond the confines of these conventional provinces, 
or beyond a linguistic stock that more or less coincides with a 
cultural group, the fact of transmission is emphasized by the type 
of distribution found. Thus the shooting of a magical object with 
intent to stun candidates for initiation into the Midewiwin Society 
occurs among the Central Algonkian. In one form or another, this 
shooting is also a feature of societies among several Siouan tribes; 
but these are precisely those tribes which have been in close con- 
tact with the Central Algonkian — the eastern Dakota, southern 
Siouan, and Winnebago. The Sun Dance offers another case in 
point. This ceremony is found among the majority of Plains 
tribes, but has also been celebrated by several divisions of the 
Shoshonean stock, who properly belong, not to the Plains, but to 
the Plateau area. Here, again, the type of distribution is such as 
might be expected on the theory of diffusion: of the Shoshoni 
proper, the Lemhi did not practise the Sun Dance, but it is still 
performed at Wind River and Fort Hall, where the Shoshoni come 
more in contact with Plains peoples. 

The fact of diffusion must, then, be regarded as established; 
and the very great extent to which ceremonials have traveled 
from tribe to tribe, coupled with undoubted diffusion of other 
cultural elements in North America, indicates that, while the 
process has been greatly accelerated by improved methods of 
transportation and other circumstances promoting intertribal in- 
tercourse, it must have been active prior to these modern con- 
ditions due to white influence. 

The next problem is: How have ceremonial features been dif- 


Theories and Theorists 

fused? Plausible answers to this question seem relatively easy. 
Ceremonial regalia were often carried in war, and might readily 
be imitated, or snatched away from the enemy, and thus become a 
ceremonial feature of a new tribe. Among the Kwakiutl and 
their cognates, alien dance regalia were often secured by killing the 
owner.^^ During meetings of friendly tribes, dances were some- 
times performed for the entertainment of the visitors, who might 
thus learn a new ceremony. It was in this way that the River 
Crows came to have their Muddy Mouth performance.^*^ Where- 
ever a ceremony was considered (as frequently happened) a 
form of property, the right to perform it was naturally transferable 
to an alien who paid the customary amount of goods. Thus the 
Hidatsa secured the Hot Dance from the Arikara by purchase. 

Before going further, we must be clear as to what is really 
transmitted through the agencies suggested. For example, the 
method of acquiring certain regalia through killing the owner 
does not account for the diffusion of the ceremony itself which 
these regalia symbolize. Take an instance cited by Boas. The 
Matilpe had not been permitted by the other tribes to acquire the 
Cannibal performer's regalia. At one time their village was ap- 
proached by a party of men and women from the northern tribes, 
one of the men wearing the badge of the Cannibal order. Two 
Matilpe youths killed the strangers, and one of them assumed 
the Cannibal's cedar-bark ornaments, and at once began to utter 
the characteristic Cannibal cry, "for now he had the right to use 
the dance owned by the man whom he had killed." It is clear 
that the knowledge of the performance preceded the acquisition 
of the badge. In the native mind, to be sure, the Cannibal Dance 
was a form of property that could be acquired by killing the 
owner; and before its acquisition it did not, from the native point 
of view, form part of the Matilpe culture. But in reahty, of course, 
it did form part of that culture; for otherwise the attitude of the 
Matilpe, both before and after the murder, would be impossible. 
The essential problem involved is not how the Matilpe secured 
the symbols of the ceremony (however important these may ap- 

^ Boas, "The Social Organization ... of the Kwakiutl Indians," pp. 424—431. 
^' Lowie, Societies of the Crow . . . Indians, pp. 197 S. 


Ceremonialism in North America 

pear to the native mind), but how the Matilpe came to par- 
ticipate in the knowledge of the ceremonial. The murder did not 
effect simple bodily introduction of a new ceremony, but only 
bodily introduction of new ceremonial badges, which were fitted 
into their customary ceremonial associations through prior knowl- 
edge of the ceremonial complex to which they belong. 

It is, however, quite intelligible how such knowledge spread 
to the Matilpe through simple attendance as onlookers at per- 
formances of other tribes, for in that capacity they were hardly 
in a different position from the uninitiated spectators who be- 
longed to the tribe of the performers. Whether an observed cere- 
monial routine is actually imitated ( as in the case of the Muddy 
Mouth Dance of the River Crows), or remains unexecuted, con- 
tingent on fulfilment of requirements due to existing property 
concepts, is, from the point of view of diffusion, relatively unim- 
portant. The point is that not only tangible articles, but even an 
objective series of acts, songs, etc., may readily spread from tribe 
to tribe. In Australia it has been proved that ceremonies travel 
in various directions, like articles of exchange, and that frequently 
"a tribe will learn and sing by rote whole corrobborees in a lan- 
guage absolutely remote from its own, and not one word of 
which the audience or performers can understand the meaning 
of." ^'^ Illustrations of similar forms of borrowing are not lacking 
in North America. Thus the Winnebago chant Sauk songs during 
their Medicine Dance; and the music of songs is readily passed on 
from tribe to tribe, as in the case of the Grass Dance. 

When there is esoteric ceremonial knowledge, the process of 
transmission implies, of course, far more intimate contact. Here 
the borrowing individuals or groups must be treated, for purposes 
of initiation, as though they belonged to the tribe from which the 
knowledge is obtained. The Arikara trick of plunging one's arm 
into scalding hot water without injury could not be imitated by 
the Hidatsa on the basis of mere observation; instruction must 
be bought, as it would be bought by an Arikara novice from an 
Arikara adept. Through similarly close personal contact, the 

^'W. E. Roth, Ethnological Studies among the North-West-Central Queensland 
Aborigines (Brisbane, 1897), p. 117. 


Theories and Theorists 

Medicine Pipe ceremony spread from individual Arikara to 
individual Hidatsa, and from individual Hidatsa to individual 

To sum up: transmission of external features, such as ceremonial 
paraphernalia, is possible on the basis of superficial, possibly even 
hostile, meetings; friendly intertribal gatherings render possible 
the borrowing of ceremonial routine, songs, and the like, in short, 
of the exoteric phases of the complex; while initiation into the 
inner meaning of a ceremony becomes feasible only through the 
closest form of personal contact. 

Nevertheless the problem of diffusion is still far from being ex- 
hausted. Even where a ceremony seems to be bodily transferred, 
it may become different because of the differences in culture be- 
tween the borrowing and transmitting tribes; that is to say, even 
an entire ceremony is not an isolated unit within the culture of 
the tribe performing it, but has definite relations to other cere- 
monies and to the tribal culture generally. Even tribes sharing 
in large measure the same mode of life tend to diverge as regards 
specific conceptions of social and ceremonial procedure. The 
"same" ceremony may thus enter different associations, and in so 
far forth become different through its novel relations. There can 
be no doubt that the Tlingit and Haida potlatches represent a 
single cultural phenomenon. Nevertheless there is a remarkable 
disparity between the associations of the great potlatches of these 
tribes. Among the Haida, the main festival was conducted by a 
chief in behalf of his own moiety, and was intended only to en- 
hance his social standing. The Tlingit performed a potlatch for 
the benefit of the complementary moiety and for the sole avowed 
purpose of showing respect for the dead.^^ This illustration is in- 
structive, because it embodies both types of changes that a trans- 
mitted ceremony undergoes, — a change in objective relations, 
which, however, cannot in many instances fail to affect the sub- 
jective attitude of the performers or borrowing tribe at large; 
and a change of the ostensible object, of the theoretical raison 
d'etre, of the performance. These types of changes had best be 
considered separately. I shall approach the primarily objective 

^^ Swanton, "Social Condition ... of the Tlingit Indians," pp. 434 ff.; idem, 
Contribution to the Ethnology of the Haida, pp. 155 ff., 162. 


Ceremonialism in North America 

alterations undergone by a borrowed ceremony through a con- 
sideration of the specific tribal patterns for ceremonial activity; 
and I shall consider the changes of avowed raison d'etre in 
diffused ceremonies in the section dealing in a general way with 
the ends sought through ceremonial performances. 

To avoid misunderstanding, it must be noted that by no means 
all changes of diffused ceremonies can be brought under these 
two heads. This is best seen when comparing the established 
variations in the performance of the same ceremony by local 
subdivisions of the same tribe. Thus we find that in some Haida 
towns the Grizzly Bear spirit inspired only women, while in 
others there was no such restriction.^^ The River Crows adopted 
the Crazy Dog Dance from the Hidatsa without assimilating it to 
the old Crow dances, while the Mountain Crows at once as- 
similated it to the rivalry concept of their Fox and Lumpwood 
organizations. ^° The unique historical conditions upon which such 
changes of borrowed ceremonies depend are not different in type 
from those which determine modifications in an indigenous cere- 
mony, and are in neither case amenable to generalized treatment. 

Among the Arapaho the seven ceremonies distinctive of the age- 
societies, as well as the Sun Dance, are performed only as the re- 
sult of a pledge made to avert danger or death.^^ The dances of 
the Kwakiutl, differing in other respects, resemble one another 
in the turns about the fireplace made by entering dancers; para- 
phernalia of essentially similar type (head-rings, neck-rings, 
masks, whistles) figure in Kwakiutl performances otherwise dis- 
tinct; and the object of apparently every Kwakiutl society's winter 
ceremonial is "to bring back the youth who is supposed to stay 
with the supernatural being who is the protector of his society, 
and then, when he has returned in a state of ecstasy, to exorcise 
the spirit which possesses him and to restore him from his holy 
madness." '*^ Among the Hidatsa the right to each of a considerable 
number of esoteric rituals must be bought from one's father: in 

™ Swanton, Contribution to the Ethnologij of the Haida, p. 171. 

*" Lowie, Societies of the Crow . . . Indians, p. 148. 

" Kroeber, The Arapaho, pp. 158, 196. 

*^ Boas, "The Social Organization ... of the Kwakiutl Indians," pp. 43 ff. 


Theories and Theorists 

each case the requisite rituahstic articles were supphed by a clans- 
man of the buyer's father; a "singer" conducted the ceremonies; 
the purchaser received the ceremonial bundle, not directly, but 
through his wife; and so forth.^^ All important bundle ceremonies 
of the Blackfeet require a sweat-lodge performance; in nearly all 
rituals the songs are sung by sevens; for almost every bundle some 
vegetable is burned on a special altar; and every ritual consists 
essentially of a narrative of its origin, one or more songs, the 
opening of the bundle, and dancing, praying, and singing over its 
contents. ^^ 

It would be manifestly absurd to assume that the notion of 
performing ceremonies to ward off death originated eight times 
independently among the Arapaho; that the originators of the 
Kwakiutl Cannibal ceremonial and the originators of the Kwakiutl 
Ghost Dance independently conceived the notion of wearing neck- 
rings; ^^ and so forth. Wissler has forcibly brought out the point 
that among the Blackfeet the Beaver Bundle owners seem to have 
established a pattern of ceremonial routine that has been copied 
by the owners of other bundles; and many additional illustrations 
could be cited to prove that, in every tribe with a highly developed 
ceremonial system, a corresponding pattern has developed. The 
psychology of this development has been felicitously compared 
by Goldenweiser with the process of borrowing ideas from an 
alien tribe: in both cases a novel idea is suggested, and may be re- 
jected, or partly or wholly assimilated.^^ Whenever such an idea 
is generally adopted within a tribe, it tends to assume the char- 
acter of a norm that determines and restricts subsequent thought 
and conduct. The Plains Indian generally ascribes any unusual 
achievement, not to personal merit, but to the blessing of a su- 
pernatural visitant; hence he interprets the invention of the phono- 
graph in accordance with this norm. Among the Hidatsa it is cus- 
tomary to give presents to a father's clansman; hence an Hidatsa 

*' My own field notes. 

" Wissler, Ceremonial Bundles, pp. 257, 271, 254, 101, 251. 

*' Boas, "The Social Organization ... of the Kwakiutl Indians," in wliich cf. 
figs. 81, 147. 

" A. A. Goldweiser, "The Principle of Limited Possibilities in the Development 
of Culture," Journal of American Folk-Lore, XXVI (1913), 287. 


Ceremonialism in North America 

purchasing admission into an age-society selected from among the 
group of sellers a member of his father's clan. The notion at the 
bottom of the norm originates, of course, not as the notion of a 
norm, but like all other thoughts that arise in individual con- 
sciousness; its adoption by other members of the social group is 
what creates the pattern. We cannot, without tautology, generalize 
as to the type of concept that will become a model; indeed, we 
have found that, in two difFerent bands of the same tribe, an 
already established concept may in the one case assimilate an 
alien introduction, and in the other capriciously fail to exert any 
influence on it. All that we can say is, that patterns exist, and are 
one of the most active forces in shaping specific cultures. 

From the point of view here assumed, a problem that might 
otherwise arise in the study of North American ceremonialism, and 
has already been touched upon, assumes a somewhat different 
aspect. Finding a very complex ceremonial system in certain parts 
of the continent, in the absence of such a system in others we 
might be tempted to ascribe the difference to a psychological dif- 
ference between the respective tribes. In some measure, to be 
sure, extensive diffusion of cultural elements in some areas as 
compared with others would account for the observed phe- 
nomenon. If at one time the tribes of the Northwest coast or 
the Plains, taken singly, possessed a ceremonial culture as simple 
as that of California or the Plateaus, but spread their respective 
ceremonials among other tribes of the same area whose cere- 
monials they in turn adopted, then complexity might ensue with- 
out any cause other than conditions favorable for cultural dis- 
semination. On the other hand, the purely internal action of the 
pattern principle would sujBBce to produce a corresponding com- 
plexity. The Crows have a Tobacco order composed in recent 
decades of perhaps a dozen or more distinct branches or societies, 
all sharing the right to plant sacred tobacco, and differing only 
in the specific regalia, and instructions imparted to the founders 
in the visions of other experiences from which the branches are 
derived. Visions of similar type are not lacking among such a tribe 
as the Shoshoni; but in the absence of an integrating pattern they 
have not become assimilated to a ceremonial norm. A Crow who 


Theories and Theorists 

belonged to the Tobacco order, and stumbled across a nest of 
curiously shaped eggs, would form an Egg chapter of the Tobacco 
order; a Shoshoni might experience precisely the same thrill un- 
der like conditions, but the same psychological experience could 
not possibly result in the same cultural epiphenomenon. The 
several Tobacco societies of the Crow do not represent so many 
original ideas, but are merely variations of the same theme. There 
is, then, only one basic idea that the Crow have and the Shoshoni 
have not, — the idea of an organization exercising certain cere- 
monial prerogatives, for the ceremonial features in themselves 
are of a type probably not foreign to any North American group. 
The complexity of the socio-ceremonial life of the Crows is thus 
an illusion due to the fact that this single idea became a pattern. 
The pattern principle is also of the greatest value in illuminat- 
ing the precise happenings during the process of diffusion. It 
has been shown in another section, that a borrowed ceremony, 
even when bodily adopted, becomes difiFerent, because it origi- 
nally bore definite relations to other cultural features of the 
transmitting tribe; and, unless these additional features happen 
to exist in the borrowing group, the same unit must assume a dif- 
ferent cultural fringe. What happens in many, perhaps in the 
majority of, such cases, is that the borrowed elements are fitted 
into conformity with the pattern of the borrowing tribe. Thus the 
Dog Society of the Crows is traced back to the Hidatsa. But among 
the Hidatsa this ceremonial body is one of a graded series of 
military societies in which it occupies a definite position; and 
entrance into it, as in the case of the rest, is a matter of purchase. 
Since the Crows neither grade their military organizations nor 
exact an entrance fee in any of them, the Dog Society naturally 
lost the impress of the Hidatsa mold so far as these features were 
concerned. Moreover, it was made over to fit the Crow scheme. 
Entrance into the society was, as for all other Crow military 
societies, either a matter of choice, or, more commonly, was 
stimulated by the desire of members to have the place of a de- 
ceased member filled by a relative. Again, while police duties 
among the Hidatsa were the exclusive right of the Blackmouth 
Society, the Crow organizations all took turns at exercising this 
social function, the Dog Society among the rest. Thus the Dog 


Ceremonialism in North America 

Society with all its ceremonial correlates came to enter quite new 
combinations and to assume a specifically Crow aspect. ^^ 

To Radin we are indebted for a suggestive investigation of the 
mechanism of ceremonial borrowing with special reference to the 
selective and assimilative influences exerted by the recipient cul- 
ture on the borrowed features. The peyote cult, a very recent im- 
portation from Oklahoma, has rapidly risen to a most important 
position in the life of the Nebraska Winnebago. A detailed study 
indicates that the only really new thing introduced was the peyote 
itself, its ceremonial eating, and its effects. Several Christian ele- 
ments that enter into the present Winnebago performance prove 
to be similar to preexisting aboriginal concepts, so as to suggest 
that their acceptance was due to this conformity. The founder of 
the Winnebago cult seems to have at once placed the new plant 
in the category of medicinal herbs, and accordingly to have as- 
sociated with it the traditional shamanistic ideas. The organiza- 
tion of the new society automatically conformed to the Winne- 
bago norm. The origin narrative developed by one of the converts 
"assumed all the characteristics of a Winnebago fasting experi- 
ence and ritualistic myth, similar to those connected with the 
founders of the old Winnebago cult societies. In its totality, the 
atmosphere of the peyote cult became thus liighly charged with 
the old Winnebago background." ^^ 

Speaking of the Mandan Okipa, Catlin recognizes three "dis- 
tinct and ostensible objects for which it was held": it was an 
annual commemoration of the subsidence of the deluge; it was 
an occasion for the performance of the Bull Dance, which caused 
the coming of buffalo herds; and it was conducted in order to 
inure young men to physical hardship, and enable the specta- 
tors to judge of their hardihood.^^ The diversity of these alleged 
objects suffices of itself to suggest that the Okipa is a complex 
performance; that it would be vain to try to account for its ori- 

" Lowie, "Some Problems in the Ethnology of the Crow and Village Indians," 
American Anthropologist, XIV (1912), 70; idem, Societies of the Crow . . . 
Indians, p. 155. 

** Radin, "A Sketch of the Peyote Cult of the Winnebago." 

"George Cathn, 0-kee-pa (London, 1867), p. 9. 


Theories and Theorists 

gin by a simple psychological explanation. It is a priori psycho- 
logically conceivable that the Okipa (that is, an annual four- 
day summer festival) originated as a celebration commemora- 
tive of the mythical flood, how^ever improbable this may appear 
from our considerations of "Myth and Ritual"; but, if so, the con- 
ception that it was intended to attract the buffalo and the concep- 
tion that it was an ordeal for the young men were secondary. Or 
we may assume that the ordeal concept was primary; then the two 
other alleged functions were secondary. And a corresponding 
conclusion seems inevitable if we suppose that the enticing of 
the buffalo was the original motive for the festival. In a more ac- 
ceptable form, this theory might be stated as assuming that three 
originally independent ceremonies performed for diverse ends 
somehow became welded together into what then became the 

Before going further, it will be well to demonstrate that the 
complexity of the ceremony is an historical fact. This becomes at 
once obvious when we consider the distribution of two of our 
three hypothetical elements. The buffalo-calling ceremony is by 
no means a peculiarity of the Mandan Okipa, but a ceremony 
very widely diffused over the Plains area: indeed, a buffalo-call- 
ing ceremony not differing in principle from that of the Okipa 
was performed by the Mandan themselves independently of the 
Okipa; ^^ and a ceremony undertaken for the same ostensible 
purpose and with corresponding mimetic features was practised 
by the Mandan White Buffalo Cow Society.^^ What is true of the 
buffalo-calling feature applies with even greater force to the volun- 
tary self-torture element. This appears with all its characteris- 
tic details — such as piercing of the breasts, insertion of skewers, 
suspension from a pole, and dragging of buffalo-skulls — not only 
in the Sun Dance of various tribes (where there is a collective 
torture strictly comparable to that of the Okipa), but also among 
the Dakota, Crows, and other Plains peoples, as a fairly normal 
procedure in the individual quest for supernatural aid.''" That the 
buffalo-calling ceremony and the specific self-torturing practices 

Maximilian, op. cit., II, 181, 264 ff. 

Lowie, Societies of the Crow . . . Indians, pp. 346-354. 

J. O. Dorsey, op. cit., pp. 436 ff. 


Ceremonialism in North America 

under discussion were at one time independent of each other, and 
of whatever other features they are combined with in the Okipa, 
must be considered an estabHshed fact: indeed, the complexity is 
greater than the theory here discussed would indicate. To men- 
tion but one conspicuous feature, a great deal of time is consumed 
in the Okipa with dances by mummers impersonating animals 
and closely mimicking their appearance and actions. The per- 
formances are objectively, in a rough way, comparable to the 
Bull Dance, but have nothing to do with any solicitude for the 
food supply, since many of the beings represented are not game 
animals. These animal dances rather suggest the dream-cult cele- 
brations of the Dakota, especially as the performers chanted 
sacred songs distinctive of their parts, and taught only on initia- 
tion and payment of heavy fees.^^ The mimetic animal dance thus 
forms an additional element of the Okipa complex. 

The complex character of the ceremony is thus an historical 
fact. How, then, shall we interpret the equally certain fact that, 
to the native consciousness, it appeared as a unified performance 
instituted by the mythical hero Nii'mak-maxana,^^ and celebrated, 
if not for the specific reasons assigned by Catlin, from the vaguer 
motive of promoting the tribal welfare in general? ^^ 

We shall not go far wrong in putting the alleged raison d'etre 
of the Okipa in the same psychological category with ritualistic 
myths. As the myth is an aetiological afterthought associated 
with a preexisting rite, so the alleged object of a complex cere- 
mony may be merely an afterthought engrafted on a preexisting 
aggregation of ceremonial elements. In the one case it is the 
aetiological, in the other the teleological, feature that welds to- 
gether disparate units, and creates the illusion of a synthetized 
articulated whole. If the hero Nii'mak-maxana ordered the Man- 
dan to practise a particular combination of un-unified observ- 
ances, these performances become unified by that mythical fiat; 
and the causal requirements of the native, at the stage when 
rationalization sets in, are satisfied. At this stage the teleological 
point of view naturally serves the same purpose: in practice, in 

^' Catlin, op. cit., pp. 19 ff.; Maximilian, op. cit., II, 178. 

^ Maximilian, op. cit., II, 172. 

"^E. S. Curtis, The North American Indian (Cambridge, 1907-1930), V, 26. 


Theories and Theorists 

fact, it largely coincides with the aetiological attitude. If Nii'mak- 
maxana instituted the annual festival, he did so for the purpose 
of benefiting the Mandan, and dereliction would spell tribal 
disaster. On the other hand, if the ceremony insures the common- 
weal, no further cause for its performance is required. 

The principle here illustrated by the Okipa may be demon- 
strated in even more satisfactory fashion for the Sun Dance of 
the Plains tribes. Whatever may be the avowed purpose of this 
performance, certain elements are practially uniform throughout 
the area; for example, the selection and felling of a tree treated 
as an enemy, the erection of a preparatory and a main lodge, and 
a several-days' fast culminating (except among the Kiowa) in 
torture proceedings of the Okipa type. The Sun Dance of the 
Crows was performed exclusively in order to secure vengeance 
for the slaying of a tribesman; among the western Algonquian 
tribes it was vowed in the hope of delivering the pledger or liis 
family from sickness or danger; wliile benefits of a vaguer and 
more public character were expected by the western Dakota, 
Hidatsa, and Kiowa.^^ In view of this diversity of ends sought, 
we cannot associate the ceremonial routine defined above with 
any of the ostensible objects of the Sun Dance; for in all cases 
but one the object must be secondary, and, from an argument 
analogous to that used in the consideration of "Myth and Ritual,' 
the residual case appears amenable to the same psychological in- 
terpretation. In other words, the ostensible motive of complex 
ceremonies is not the genuine or original motive, but embodies 
merely the present native theory of the reason for the performance. 

Several questions naturally arise: If we cannot directly in- 
terpret a complex ceremony, can we not at least give a psychologi- 
cal interpretation of its components? further, if we can resolve 
it into such constituents, how must we conceive the process by 
which originally unrelated elements became joined together (as 
we have assumed) through historical accident, to be integrated 
only at a later stage by some rationalistic synthesis? and, finally, 

^'' G. A. Dorsey, The Arapaho Sun Dance, pp. 5 ff.; idem. The Cheyenne, p. 58; 
Walter McClintock, The Old North Trail (London, 1810), p. 170; Kroeber, 
Ethnology of the Gros Ventre, p. 251; H. L. Scott, "Notes on the Kado, or Sun 
Dance of the Kiowa," American Anthropologist, XIII (1911), 347; J. O. Dorsey, 
op. cit., p. 451. 


Ceremonialism in North America 

if the native theory is merely an interesting speculative misinter- 
pretation of native psychology, what is the present psychological 
correlate of those complicated series of observances under dis- 

Let us consider first of all the second question. Analysis re- 
solves a ceremony into a number of disparate elements; how did 
these ever become joined together? We are here confronted by 
the problem of secondary association, a large topic to which only 
a few words can be devoted in this article. In the first place, we 
should beware of confounding logical with historical analysis. 
Two features may be not only logically as distinct as musical 
pitch and timbre, but also as inseparable in reality. This principle 
has already been expressed by Dr. Radin, though his illustration 
rather shows how apparently unrelated concepts are nevertheless 
logically related in the native mind. The notion of a society de- 
rived from a water-spirit and the notion of curing disease are 
apparently distinct; but, if the water-spirit is always associated 
with the granting of medical knowledge, a vision of the water- 
spirit and the acquisition of medical skill coincide. Thus, what- 
ever may be the development of the conception entertained re- 
garding the water-spirit, the association between the idea of a 
society based on a supernatural communication by that spirit 
and the idea of doctoring is primary.^^ Here the initial disparity 
of the elements found in combination proves to be apparent, 
being merely due to our ignorance of the tertitim quid. A pri- 
mary ceremonial ^^ association of genuinely distinct and cere- 
monially indifferent objects may be achieved through their juxta- 
position in a vision, as illustrated by many medicine bundles. 
Thus, a jackrabbit-skin and a bunch of eagle-feathers may together 
form an ultimate unit of ceremonial stock-in-trade. 

Let us now turn to cases of association of elements once exist- 
ing apart. One cause of secondary association has already been 
touched upon. Wherever a particular ceremonial concept be- 
comes the predominant one, it tends to assimilate all sorts of 

^^ Radin, "The Ritual ... of the Winnebago Medicine Dance," pp. 193, 196. 
The point seems to me to be closely related to that repeatedly made by Le\'i-Baihl 
in his Les Fonctions mentales dans les societes inferieures (Paris, 1922), with 
reference to "participation." 

^ Otherwise, of course, the association is secondary. 


Theories and Theorists 

other concepts originally independent of it: thus, in the Crow 
example of the Tobacco societies and in the case of the Blackfeet 
Beaver Bundle, which has not only become the pattern for other 
bundles, but has even absorbed such rituals as the Sun Dance and 
Tobacco ceremony.^^ Among the Crows, individual visions by 
members of the Tobacco order have led to the association of 
quite heterogeneous features. A Tobacco member who chanced 
upon curiously-shaped eggs would found an Egg chapter of the 
order, and initiate new members into it, thus bringing about a 
connection between egg medicine and the sacred Tobacco; and 
in corresponding fashion have developed the Weasel, Otter, Straw- 
berry, and other divisions. 

In these cases it would seem that the notion of sacredness or 
ceremonialism is so strongly associated with a particular content 
that has become the ceremonial pattern that any new experience 
of corresponding character is not merely brought under the same 
category as the pattern, but becomes an illustration, an adjunct 
of the pattern concept. In many other instances, a ceremony may 
bring about conditions normally associated with certain activities 
in no way connected with the ceremony itself; and, when these 
conditions arise in the course of the ceremony, they act as a 
cue to the performance of the normally associated activities. 
There is no connection between initiation into a society privileged 
to plant tobacco for the tribal welfare and the recounting of an 
individual's war-record; nevertheless, in the Crow Tobacco adop- 
tion, the entrance into the adoption lodge is uniformly foUowed 
by such a recital. The reason is fairly clear. At every festive gather- 
ing of the Crows there is a recital of war-deeds; the Tobacco in- 
itiation produces such a gathering, which elicits the customary 
concomitant; and thus the coup-recital becomes a feature of the 
Tobacco adoption ceremony. Similarly, every Iroquois festival 
seems to have been preceded by a general confession of sins.^** 
Still another way by which heterogeneous ceremonial activities 
or features become associated is, of course, by purchase. The 
Hidatsa Stone-Hammer Society, according to Maximilian, bought 
the Hot Dance from the Arikara. But the Stone-Hammers had a 

^' Wissler, Ceremonial Bundles, p. 220. 
^ L. H. Morgan, op. cit., p. 187. 


Ceremonialism in North America 

ceremony of tlieir own prior to the purchase, which was thus as- 
sociated with the newly acquired fire-dance and the plunging of 
arms into hot water. 

These few suggestions must suffice to indicate how disparate 
elements may become secondarily associated. 

So far as the interpretation of the single elements is concerned, 
there is relatively little difficulty. Though we may not be able 
to comprehend the ultimate origin of a certain mode of cere- 
monial behavior, we can generally apperceive it as typical of a 
certain tribe or a certain group of tribes. The fact that the Plains 
Indians went to fast in a lonely place, looking for a supernatural 
revelation, may remain an irreducible datum; but, when we dis- 
engage from the Crow Sun Dance complex the attempt to secure 
a vision that is given as its ultimate motive, we at once bring 
it under the familiar heading of "vision-quest." So we may not 
know how "four" came to be the mystic number of many tribes; 
but it is intelligible that, where it is the mystic number, dances, 
songs, processions, and what not, should figure in sets of four. 
Prayers, dances, sleight-of-hand performances, the practice of 
sympathetic or imitative magic, etc., are likewise ultimate facts; 
but their special forms in ceremonies of which they are part are 
readily classified with corresponding psychological manifesta- 

But the social setting of the cultural elements enumerated dur- 
ing a ceremony cannot fail to lend them a color they otherwise 
lack. The pledger of the Crow Sun Dance, who sets in motion 
the tremendous machinery required for the communal undertak- 
ing, and is thenceforth subjected to tribal scrutiny, cannot be 
supposed to be in the same psychological condition as if he were 
merely seeking a vision in the seclusion of a four-night vigil on a 
mountaintop. What we find in any complex performance of this 
type, then, is a number of distinct acts with distinct psychological 
correlates, integrated, not by any rational bond, but by the cere- 
monial atmosphere that colors them all. 

From this point of view the question "What may be the object 
or psychological foundation of a ceremony?" becomes meaning- 
less. The psychological attitude is not uniform for tlie performers 
of a ceremony: it is not the same for the Sun Dance pledger (who 


Theories and Theorists 

wishes to compass an enemy's death) and the self-torturing vi- 
sion-seekers in quest of martial glory. Much less is it the same 
for the pledger and the self-advertising reciters and enactors 
of war-exploits or the philandering couples hauling the lodge- 
poles. But is not the attitude of the pledger the essential thing? 
to assume this customary view is the surest way to miss the 
nature of ceremonialism. A Crow Sun Dance pledger wishes 
to effect the death of an enemy; a Cheyenne Sun Dance pledger 
wishes to insure the recovery of a sick relative. Why must both 
have, say, a dramatic onslaught on a tree symbolizing an enemy? 
From the rationalistic point of view here criticized, the answer 
is not obvious. It would be in perfect accord with the Plains In- 
dian mode of action for the Crow and Cheyenne simply to retire 
into solitude and secure a vision bringing about the desired re- 
sult. If they are not content with this, and require an elaborate 
ceremonial procedure, that procedure must have an additional 
raison d'etre. The absence of intelligible object (from the native 
rationalistic point of view no less than from our own) in a cere- 
monial feature becomes at once clear, if we regard its very per- 
formance as self-sufficient, as gratifying certain specific non- 
utilitarian demands of the community. View it not as primitive 
religion, or as a primitive attempt to coerce the forces of nature, 
but as a free show, and the mystification ceases: ceremonialism 
is recognized as existing for ceremonialism's sake. 



Edward B. Tylor 

Edward B. Tylor, who died on January 2, 1917, at the 
age of eighty-four, had long been an historic personahty. He 
loomed up as one of the very last figures rooted in the heroic age 
of nineteenth-century science, as the peer and comrade in arms of 
Wallace, Huxley, and Spencer. The dean of ethnologists for two 
score years, he represented his science before students of other 
branches of knowledge and, thanks to the high literary quality 
of his style, before the cultured laity as well. He was read and 
cited by psychologist and historian, biologist and philosopher, by 
everyone interested in the ways and thoughts of primitive man. 
And while the circle of his influence widened, he retained the 
profound and growing respect of his professional colleagues. 
Even with the irreverent group of American field workers who 
turn up their noses at the classical school of ethnologists his 
prestige remains undiminished and their allegiance is of the kind 
he himself advocated, — no slavish acceptance of tenets but a 
following of methods "through better evidence to higher ends." 
Edward Burnett Tylor was born at Camberwell on October 
2, 1832, and educated at Grove House School, Tottenham. After 
a brief business career he traveled for several years and in 1856 

American Anthropologist, XIX (April-June, 1917), 262-268. 


Theories and Theorists 

visited Mexico in the company of Henry Christy, an anthropologist 
to whose personal stimulation he pays a generous tribute in the 
second edition of the Researches. The American trip led to Tylor's 
first publication, a book on Anahuac; or Mexico and the Mexicans 
( 1861 ) . Several years later appeared the Researches into the Early 
History of Mankind and the Development of Civilization (1865). 
This work laid the foundation of his professional fame, which 
reached its acme in 1871 with the publication of Primitive Culture: 
Researches into the Development of Mythology, Philosophy, Re- 
ligion, Language, Art, and Custom. In 1881 he wrote a most serv- 
iceable textbook on Anthropology: An Introduction to the Study of 
Man and Civilization. 

Though not a university graduate, Tylor became connected 
with Oxford, both in the capacity of keeper of the University 
Museum and as a lecturer, being "reader in anthropology" from 
1884 to 1895 and "professor" from 1895 to 1909, when he became 
an emeritus. Of the numerous honors conferred on him only two 
need be mentioned here. He was elected to a fellowship by the 
Royal Society in 1871 and knighted in 1912. A volume of Anthro- 
pological Essays Presented to Edward Burnett Tylor in Honor 
of His Seventy-fifth Birthday bore testimony to the regard of his 
fellow- workers. The bibliography concluding that volume indicates 
the extraordinary number of smaller and scattered contributions 
that fell from his pen in the course of years, and we learn with 
deep regret that a great work he had been preparing for many 
years was never published, which was also the fate of his ten 
Gilford lectures on Natural Religion, delivered at Aberdeen in 

The most obvious feature that distinguishes Tylor's work from 
that of his English contemporaries and successors is the univer- 
sality of his ethnological interests. Others, like Lang and Frazer, 
were predominantly occupied with sociological and religious 
problems; Tylor's vision embraced, to cite his own definition of 
culture, "that complex whole which includes knowledge, belief, 
art, morals, law, custom, and any other capabilities and habits 

^ The biographical data are taken from Lang's sketch in the anniversary vokime 
cited above and from Professor Alfred C. Haddon's obituary notice in Nature, 
January 11, 1917, p. 373. 


Edward B. Tylor 

acquired by man as a member of society." He was equally at- 
tracted by the description of a Malagasy bellows and by an ac- 
count of the South American couvade, by the process of stone- 
boiling and by solar mythology. 

In Tylor's attitude towards the immense mass of concrete fact 
with which his versatility brought him into contact a distinctive 
psychological trait is manifest — his intuitive sense of fitness. We 
must recall the character of the data available when he com- 
menced his life work — the hodge-podge of imperfect observation 
and provincial bias with which he was obliged to deal in order 
to get at the mere facts. To be sure, there was excellent material 
by men like Cranz, Sahagun, or Callaway. But even the most 
reputable of the older writers were prone to state as fact what 
was either crude misinformation at second-hand or crude mis- 
interpretation due to the colored spectacles of European civiliza- 
tion. What shall we say when we find Burton declaring that the 
Arapaho possessed so scanty a vocabulary that they could hardly 
converse with one another in the dark when gestures were in- 
visible, or Baker denying any form of religion to the aborigines 
of the Upper Nile region? In the evaluation of such utterances 
Tylor showed an almost unerring instinct, all the more commend- 
able since many of the wild statements of this type would have 
fitted admirably into that general evolutionary scheme of the 
universe which he himself was helping to develop. 

This critical judgment was apparent in the discussion of prob- 
lems as well as in the weighing of travelers' accounts, but here 
the result was not so uniformly satisfactory. Indeed, the question 
obtrudes itself, whether Tylor's famous caution was not some- 
times conformity to a scientific ethical ideal of fairness in discus- 
sion rather than a trait inherent in his mental make-up. He cer- 
tainly carried the judicial weighing of pros and cons to an 
exceptional degree. On rereading the Researches into the Early 
History of Mankind, I can understand Wallace's irritation at its 
indecisiveness and Lubbock's misunderstanding of the argument 
as to the single origin or independent development of the couvade. 
But whatever formal hedging there may be in the marshaling of 
arguments, the conclusion sometimes appears as a thunderbolt 
out of a blue sky, as when historical connection is used to interpret 


Theories and Theorists 

the existence in remote areas of the cure by extracting pathogenic 
agents from the patient's body. 

This illustration, however, brings up a topic which shows Tylor 
to the greatest possible advantage in historical perspective. Though 
certainly a strong believer in the independent evolution of cul- 
tural phenomena in distinct areas of the globe, he was very 
much alive to the influence of diffusion. In the Introduction to 
the Enghsh translation of Ratzel's History of Mankind he con- 
trasts "the small part of art and custom which any people may 
have invented or adapted for themselves" with "the large part 
which has been acquired by adopting from foreigners whatever 
was seen to suit their own circumstances." Indeed, in many con- 
crete instances he goes much further than at all events modem 
American etlmologists are inclined to follow. The case of cure by 
suction has already been cited, while another chapter of the same 
book pre-figures in principle the recent hypothesis of a cultural 
connection between aboriginal America and the Old World. What- 
ever we may think of particular interpretations offered by Tylor, 
^. the traditional American conception of him as merely an evolu- 

yy/ tionist of the classical school is ridiculously false. His suggestive 

and indeed conclusive discussion of the Malagasy iron technique 
alone suflSces to show what a valuable tool he sometimes made of 
the principle of historical connection. 

Nevertheless, it remains true that Tylor's name will always be 
most prominently connected with the doctrine of evolution. In 
this context it is very cheap to assume an unhistorically critical 
attitude. We must recollect that just as he had to sift the chaotic 
mass of ethnographic observations in order to extract the actual 
facts so in the interpretation of culture history he had to contend 
with a powerful, theologically inspired theory of degeneration 
against which the principle of progressive evolution had to be es- 
tablished and defended. To have accomplished this task so ef- 
fectively is in itself no mean achievement to Tylor's credit. But 
Tylor further enriched the doctrine of cultural evolution by the 
development of a definite and elaborate scheme for the subject of 
religion. To enter into a discussion of this theory of animism is 
out of the question within the limits of this notice. SufiBce it to 
say that as presented in Primitive Culture it remains, in spite of 


Edward B. Tylor 

all criticism, the most impressive theory of primitive religion yet 

To philosophical ethnology Tylor contributed the concept of 
survivals and the intimately associated method of "adhesions" 
outlined in his ever memorable paper "On a Method of Investigat- 
ing the Development of Institutions, Applied to Laws of Marriage 
and Descent," which was presented to the Anthropological In- 
stitute in November, 1888, and pubhshed in Vol XVIII (1889) 
of its Journal. It must be reckoned a distinct loss to science that 
the complete data on which this lecture was based were never 
published. The fundamental idea is the application of statisti- 
cal methods to the data of ethnography. If two or more cultural 
traits are repeatedly found in association, are we dealing with a 
chance combination or is there an organic correlation? Tylor com- 
pares the number of times such combinations might be expected 
to occur on the doctrine of probabilities if each feature were in- 
dependent of the others with the number of occurrences empiri- 
cally found, and where the latter is clearly in excess he infers a 
causal connection. In this manner, e.g., he establishes a functional 
relationship between the exogamous dual organization and the 
classificatory systems of kinship terminology, between the parent- 
in-law taboo and matrilocal residence, and between the couvade 
and a mixed maternal-paternal organization. 

The very idea of introducing into a branch of knowledge that 
is so often the happy hunting-ground of the curiosity-seeking dilet- 
tante something of the rigor of the exact sciences is one of well- 
nigh unparalleled magnificence. Nothing that Tylor ever did 
serves so decisively to lift him above the throng of his fellow- 
workers. Without that paper he might have ranked as a sort of 
super-Lang or super-Frazer — more universal in his grasp than 
either, more serious and erudite than the one, far more trustworthy 
in his judgment than the other. But the paper on Method raises 
him at once into an entirely different category of intellectual 

In the appraisal of this contribution several points should be 
considered separately. In the first place, quite apart from the 
main argument, Tylor here first conceptualized certain phenomena 
which have since loomed more or less prominently in ethnographic 


Theories and Theorists 

literature, viz., teknonymy and cross-cousin marriage. Secondly, he 
was fully aware of the fact that it is one thing to establish the mere 
fact that two features are causally related and quite another to 
determine the reason for the association. The former is by far the 
more important methodologically and whatever criticism may be 
advanced against Tylor's specific conception of the nature of the 
correlation does not affect the core of the method. This likewise 
remains valid even if we reject the evolutionary interpretation 
which Tylor gave to certain of his observed correlations. Finding 
no instances of the couvade among matrilineal tribes, twenty 
cases among peoples with a mixed system, and eight in patrilineal 
communities, Tylor not only inferred that the institution had 
originated in the mixed system and dwindled away with paternal 
descent but also that this established the priority of matrilineal 
descent. Obviously, this conclusion does not follow from the em- 
pirical facts of correlation but already involves the acceptance of 
a unilinear scheme of evolution. 

The essential objection to Tylor's paper, as pointed out in the 
oral discussion by Galton and Flower, rests on his neglect of dif- 
fusion. If the same combination recurs a hundred times among 
tribes that have had no historical connection, we have indeed es- 
tablished a rule of organic correlation; but if the combination has 
been disseminated from a single point of origin there is no means 
of proving that we are dealing with more than a mere chance as- 
sociation. We in America who accept difi^usion to a considerable 
extent but at the same time admit independent development 
are confronted with the fact that exactly the same usages are 
found in remote regions of the globe between which any con- 
nection remains unproved. On the other hand, these similarities 
do seem to go hand in hand with certain other similarities, with 
which therefore they seem to be functionally related. This means 
that where one of the traits occurs we can legitimately infer its 
one-time association with the correlated trait. We must insist 
against Tylor that the particular tribe in question may have bor- 
rowed the feature isolated from its old context; but to assert that 
such a correlation as that between the avunculate and a matri- 
lineal organization is due to sheer chance is ridiculous, more so 
than the wildest Graebnerism, which at least does not blink at 


Edward B. Tylor 

the observed fact of complete cultural identity. The best evi- 
dence for such an organic correlation seems to me to have been 
advanced in the field of kinship nomenclature, where Tylor him- 
self established the relation of the classificatory system with 
exogamy. But the method is applicable to an indefinite number 
of similar problems, and ethnologists will do well to turn to Ty- 
lor's extraordinarily stimulating and fruitful mode of investigation. 
Over and above his specific contributions, Tylor had a clear 
vision of the place of ethnology in modern civilization. The facts 
of primitive life were to him not mere specimens for a museum 
of psychological oddities nor was he altogether satisfied with 
using them as bricks for a theory of cultural development. Beyond 
its academic aspects he maintained that "such research has its 
practical side, as a source of power destined to influence the 
course of modern ideas and actions." The sight of mankind pain- 
fully groping through the ages from the crude fist hatchet to 
modern technology must inspire active endeavor to add to the 
heritage of the past. But ethnology also reveals in modern law, 
ethics, and theology innumerable survivals from primitive sav- 
agery, which it marks out for destruction, being in Tylor's own 
words "essentially a reformer's science." 



Lewis H. Morgan in 
Historical Perspective 

On the fiftieth anniversary of Lewis H. Morgan's 
death Russian anthropologists held a meeting in his honor; and 
The Academy of Sciences of the U.S.S.R. is publishing Ancient 
Society in its "Classics of Scientific Thought." Morgan's work 
is officially proclaimed as "of paramount importance for the 
materialistic analysis of primitive communism," while his critics 
are taunted with their bourgeois prejudices. But naturally his 
results "could not fail to fill with anxiety the hearts of those who 
connected their fate with the preservation of such relations and 
conceptions as were congenial to them." ^ 

Attempts to rehabilitate Morgan, however, are not restricted 
to Soviet philosophers. A Hindu writer dubs him "the Tylor of 
American anthropology," and a greater Tylor at that. Morgan, 
who found all primitive religions "grotesque," is quaintly credited 
with a deeper insight into religious phenomena; and still more 

Essays in Anthropology Presented to A. L. Kroeber (Berkeley: University of 
California Press, 1936), pp. 169-181. 

^ N. Matorin, "Soviet Ethnography," in Ethnography, Folklore and Archaeology 
in the USSR, IV (1933), 6; E. Kagarov, "The Ethnography of Foreign Countries 
in Soviet Science," ibid., p. 89. 


Lewis H. Morgan in Historical Perspective 

oddly, Tylor's method of "putting together disjointed scraps of 
things pertaining to man in one basket" is contrasted with Mor- 
gan's quest of "the grand harmony" in otherwise meaningless 

Morgan evidently remains more than a merely historic figure. 
It is worth while, then, to reassess his contribution and his in- 
fluence in historical perspective. The task is diflficult, but it is 
lightened by Dr. Stern's labors,^ which make it appreciably easier 
to relive the intellectual situation of Morgan's epoch. 

Morgan is profitably considered under three heads : as a gatherer 
of facts, as a philosopher of culture history, and as a contributor 
to the field of social organization. 

As an ethnographer, Morgan takes high rank. One naturally 
thinks first of his League of the Ho-de-no-sau-nee or Iroquois 
(1851), but the results of his very brief visits to more remote 
tribes are likewise most creditable. He discovered the matrilineal 
exogamous clan organization of the Crow, an observation once 
doubted but wholly confirmed by later research; and he regis- 
tered sororal polygyny as a Crow usage. Exactly as I did some 
decades later, he noted that men and women chopped off a finger 
joint in mourning or as a religious sacrifice. What is more, his 
description of the Crow kinship system is vastly superior to 
my original attempt in this direction, for he recognized that cross- 
cousins were put into different generations from the speaker's. 
As I subsequently wrote: "My error seems the less pardonable 
because the essential facts had already been grasped by Morgan." * 

Morgan's honesty as a field worker is no less conspicuous than 
his acuity. According to his theoretical scheme, the Dakota ought 
to have been organized into clans ("gentes" in his terminology); 
yet, we learn in the early 'sixties he liimseH "could find no satis- 

^ Panchanan Mitra, A History of American Anthropology (Calcutta, 1933), pp. 
109-120. The phrase from Morgan is in his Ancient Society (New York, 1877), 
Part I, chap. 1. 

' Bernhard J. Stern, Lewis H. Morgan, Social Evolutionist (Chicago, 1931); 
idem, "Selections from the Letters of Lorimer Fison and A. W. Howitt to Lewis 
Henry Morgan," American Anthropologist, XXXII (1930), 257-279, 419-453. 

* Morgan, Ancient Society, Part II, chap. 6; J. R. Swanton, "The Social Organiza- 
tion of American Tribes," American Anthropologist, VII (1905), 663-673; R. H. 
Lowie, Notes on the Social Organization and Customs of the Mandan, Hidatsa, 
and Crow Indians, AMNH, Anthropological Papers, XXI (1917), 21-26 f., 56. 


Theories and Theorists 

factory traces of gentes among them." Again, lie registers similar 
failure among the Athabaskans of Hudson's Bay territory, and 
the inability of another investigator to discover them for Morgan 
in the Slave Lake area. 

It must be admitted that Morgan was not uniformly interested 
in the whole of culture. Like most recent investigators, he devoted 
himself intensively to particular aspects of native life and skimmed 
over others. As Stern remarks, his League tells a great deal about 
Iroquois social organization, but, pace Mr. Mitra, little about re- 
ligion, and economic life is treated inadequately. However, Mor- 
gan did not neglect technology, and altogether his account re- 
mains an outstanding achievement, both for its fullness and the 
sympathy evinced toward his subjects. 

As a culture historian, Morgan was handicapped by lack of 
essential facts. Not, however, in the light of recent discoveries 
but as a contemporaneous verdict we must uphold Lubbock's 
stricture in 1878 that "Morgan's knowledge is anything but ex- 
haustive." He was indeed incomparably ahead of his times along 
special lines, yet even there his neglect of accessible sources dis- 
torted the picture he gave of ancient society. Though he had him- 
self noted the Iroquois fraternities, he never considered the 
relationship of clubs, military organizations, or religious corpora- 
tions to the social structure of American natives, let alone of 
Melanesians or Africans. Yet Maximilian's account of the Mandan 
warrior societies was available in English and German; and their 
police activities certainly had some bearing on that "Idea of 
Government" to which Morgan dedicated fifteen chapters of An- 
cient Society. 

Equally astonishing is Morgan's neglect of aristocracy and 
monarchy among iTider peoples. He laid it down as an axiom that 
monarchy was incompatible with clans, that it appeared only 
in "civilization," that is, in the period of literacy. Aristocracy, 
he argued, did not develop before the "Later Period of Bar- 
barism," that is, not before the Iron Age.^ This dogmatism hap- 
pened to yield a valuable by-product — the critical scrutiny of 
the Spanish chronicles with their extravagance about a feudal 
Aztec empire. But the general propositions were wide of the 

^ Morgan, op. cit.. Part II, chap. 5; Part IV, chap. 2. 


Lewis H. Morgan in Historical Perspective 

mark. Had Morgan never heard of the African courts described by 
early travelers? Was he ignorant of Chaka's spectacular conquests 
in the early nineteenth century? Had he ever peeped into Cap- 
tain Cook's Travels? But it was not even necessary to go so far 
afield. Caste distinctions and slavery were well-established phe- 
nomena among the natives of British Columbia. Since Morgan's 
honesty is beyond reproach, failure to note such facts can 
be imputed only to sheer ignorance. Evidently indefatigable in 
ferreting out what was relevant to clan systems and relationship 
terms, sparing no pains to acquire the kinship terminology among 
Hawaiians and Rotumans, Maori and Samoans, iie seems to have 
had no inkling of the class distinctions that prevailed in Polynesia. 
How did he interpret Tylor's statement that in eastern Asia and 
Polynesia the names of kings and chiefs were held sacred? ^ Prob- 
ably few men were further than Morgan from a "functional" view 
of diflFerent cultures as so many living wholes. 

From this deficiency springs one of Morgan's worst errors of 
classification, the inclusion of the Australians and "the greater 
part of the Polynesians" in the "Middle Status of Savagery." To be 
sure, the Australians "rank below the Polynesians," but because 
they lack bow and arrow, Maori and Kumai are both placed be- 
low the Northern Athabaskans. 

Psychologically, a flair for individual cultures and historical- 
mindedness naturally go together. Morgan lacked both; his con- 
clusions had a bearing on time sequences, but his chronology did 
not rest on archaeological stratification or written records, but 
on an abstract scheme. The peoples of the world are classified 
"according to the degree of their relative progress" into three 
periods. Savagery, Barbarism, and Civilization, the last being char- 
acterized by literacy. The two earlier periods are each subdivided 
into a Lower, Middle, and Upper Status. The Lower Status of 
Savagery was avowedly hypothetical, ending with the use of fire 
and the acquisition of a fish subsistence. The Middle Status ter- 
minated with the invention of bow and arrow; Australians and 
most Polynesians thus fall into this rubric. The Upper Status of 
Savagery ended with the manufacture of pottery, thus embrac- 

" E. B. Tylor, Researches into the Early History of Mankind and the Develop- 
ment of Civilisation (London, 1865), p. 142. 


Theories and Theorists 

ing the Indians of northern North America. Barbarism begins with 
pottery, and its Upper Status coincides with a prehterate Iron 

In the abstract, Morgan admitted his tests only with quahfica- 
tion. But in deahng with concrete material he acted precisely as 
if they were absolute. Specifically, he has sujBBcient confidence in 
them to deduce unknowable history from the scheme. This ap- 
pears precisely where up to a point Morgan did avail himself of 
written documents. So far as a general etlinologist can judge, he 
studied the sources on Greek and Roman society with pains- 
taking zeal. But while setting forth the records for the his- 
torical period, he goes far beyond ascertainable facts. According 
to him exogamy was characteristic of the Greek genos and the 
Roman gens, "a novel doctrine," as Tylor remarked in 1878, 
"which his evidence fails to establish." ''' What is more, Morgan 
admits "the absence of direct proof of ancient descent in the 
female line in the Grecian and Latin gentes," but follows Bach- 
ofen in assuming that it once existed. "It is impossible to con- 
ceive of the gens as appearing, for the first time, in any other 
than its archaic [i.e., matrilineal] form; consequently the Grecian 
gens must have been originally in this form." That is to say, 
granted a universal law of progression, that law necessarily 
applies to Greek and Roman, as well as to other cultures. If, 
however, laws give us the essence of history, why trouble at 
all about a piecemeal study of the detailed course of events in 
particular societies? Here lies the inevitable conflict of the "uni- 
linear evolutionist" and the culture historian. To the evolutionist 
it is obvious that progress has been "substantially the same in 
kind in tribes and nations inhabiting different and even dis- 
connected continents, while in the same status." Hence the 
conclusion that the archaic structure of Greek and Roman 
society "must even now be sought in the corresponding institu- 
tions of the American aborigines." The historian's approach is 
radically different. Dealing with observed sequences, he is quite 
prepared to find diverse trends among diverse peoples. When 
Laufer discusses the development of art in India, he exphcitly 
points out that it is wholly different from that of the Chinese; 

^ Quoted in Stern, op. cit., p. 141. 


Lewis H. Morgan in Historical Perspective 

while Chinese painting evolved from calligraphy, the lack of 
ornamental penmanship precluded parallelism in India, where 
the science of physiognomies played a corresponding part.^ 

Of course, all the eminent evolutionists knew that deviations 
from a norm of progress were inevitable; and Tylor expressly 
stated that most of human culture had "grown into shape out 
of such a complication of events, that the laborious piecing to- 
gether of their previous history is the only safe way of studying 
them." Far from always asserting independent development, he 
anticipates many diffusionist conclusions that are now generally 
accepted. He derived the Malagasy bellows, the Andamanese 
outrigger canoe, from Malaysia; he envisages a single origin for 
the bow and arrow; he supports the view that North American 
pottery "spread from a single source." ^ Tylor did not, as has 
been erroneously alleged, abandon this view in later years, but 
maintained it in discussing the patolli game in the anniversary 
essays in honor of Bastian. Indeed, that paper discusses the 
methodology of diflFusionism; and the only reasonable criticism 
of Tylor is that he did not uniformly balance the contradictory 
principles of explanation. Notably, his famous statistical dis- 
cussion fails to consider the effect of transmission on the course 
of organic development. 

To turn back to Morgan, he was not wholly content with the 
patter of the day. At least sporadically he asks himself what may 
be meant by similar causes: "The phrase 'similar conditions of 
society,' which has become technical, is at least extremely vague. 
It is by no means easy to conceive of two peoples in discon- 
nected areas, living in conditions precisely similar." ^" He was, 
indeed, fond of mentioning "the unequal endowments of the 
two hemispheres" as explaining the cultural differences in the 
same period, namely, the possession of domesticable animals 
that could furnish meat and milk to Old World peoples. Here, 
once more, Morgan was hampered by ignorance. It was no 

^ Berthold Laufer, Das Citralakshana, nach dem tihetischen Tanjur herau^ge- 
geben und uhersetzt (Dokumente der indischen Kunst, Heft I, Malerei; Leipzig, 
1913), p. 32; Morgan, op. cit.. Part I, chap. 1. 

« Tylor, op. cit., pp. 4, 167, 366. 

^"Morgan, Systems of Consanguinity and Affinity (Washington, 1871), pp. 
472 f.; idem, Ancient Society, Part I, chaps. 1 and 2. 


Theories and Theorists 

secret in 1877 that the same domesticated animals were quite 
diflFerently used by the Chinese and the Egyptians or the pastoral 
nomads; in other words, that "equal endowments" did not 
automatically evoke similar cultural responses. In any case, 
whatever transitory qualms may have arisen in Morgan's mind 
about the improbability of frequently similar conditions, they 
did not impress his philosophy of progress. The very sentence 
that points out the difference of the Old and the New World 
terminates with the clause: "but the condition of society in the 
corresponding status must have been, in the main, substantially 

As to diffusion, Morgan's position was strange. In his system 
the Upper Status of Barbarism begins with the Iron Age. Yet 
he puts the ancient Britons, "although familiar with the use 
of iron," into the Middle Status. "The vicinity of more advanced 
continental tribes had advanced the arts of life among them 
far beyond the state of development of their domestic institu- 
tions." As Stern remarks, this exposes the weakness of the entire 
scheme. It rests avowedly on "such . . . inventions or dis- 
coveries as will afford sufficient tests of progress," that is, on 
the arts of life. Now we suddenly learn that arts of life can 
spread by borrowing without affecting cultural status. But the 
very same principle applied to ancient Britons can be logically 
applied to an indefinite number of other cases. The vicinity of 
bow-using peoples might explain the archery of the Northern 
Athabaskans, who on other grounds might well be degraded to 
the Middle Status of Savagery, the level of Australians and 
Polynesians (sic); and so forth. 

In short, Morgan never considers what remains of a scheme 
of development if warped by incessant borrowing. Since every 
human group has been exposed to unique outside influences 
from time immemorial, how can we maintain that "the experience 
of mankind has run in nearly uniform channels?" Some manner 
of evolutionary parallelism might still be rescued from the debris, 
but surely its limitations should be recognized. 

It is no reproach to Morgan that he failed where Tylor fell 
short of complete clarity, where we of today are still flounder- 
ing. But the special ways in which Morgan invoked diffusion 
are objects of legitimate wonder. He regarded the clan ("gen- 


Lewis H. Morgan in Historical Perspective 

tile") organization as nearly universal; and from an archevolu- 
tionist convinced of a "logical progress" in human institutions 
we should expect that this would serve as a sample of what 
"the natural logic of the human mind" would everywhere produce 
in similar circumstances. Yet Morgan takes precisely the op- 
posite position: the clan is treated as so "abstruse" a conception 
that a single origin is postulated. The biological advantages of 
exogamy were such, it seems, that this type of organization 
"would propagate itself over immense areas." ^^ I am aware that 
the multiple origin of clan systems remains a moot problem; 
but nowadays the theory of a single historical source is linked 
with a general diffusionist bias, not with a belief in evolutionary 

Odd as such rank diffusionism appears in its evolutionist 
setting, it is almost mild compared with Morgan's intransigence 
in dealing with his favorite kinship data. Mere diffusion is not 
deemed sufficient here; borrowing, Morgan argues, would have 
involved the taking over of the very terms themselves. Hence, 
racial affinity alone can explain resemblances in the classifica- 
tion of relatives. Because the Hawaiian and Zulu nomenclatures 
share certain features, the Polynesians and the Kaffir must have 
sprung from the same stock; because the Tamil of India and 
the Seneca in New York state have similar systems of relation- 
ship, the affinity assumed in the name American Indian stands 
ultimately justified. ^^ 

This extraordinary intrusion of a biological factor to explain 
linguistic and sociological data cannot be treated as in any way 
due to contemporary currents of thought. Lubbock, for instance, 
at once demolished Morgan's conclusions. The Seneca resemble 
not merely the Tamil but also the Fijians and Australians in 
regard to relationship systems; are they, then, specifically related 
to all these races? Still more decisively the critic points out that 
on this basis the Two-Mountain Iroquois would have to be 
considered racially closer to the Polynesians than to the other 
Iroquois tribes. ^^ 

" Morgan, Ancient Society, Part II, chap. 15. 
^Idem, Systems, pp. 500-508. 

" Lord Avebury, The Origin of Civilisation and the Primitive Condition of Man 
(London and New York, 1911), p. 179. 


Theories and Theorists 

This case is especially illuminating, Morgan was here dealing 
with material he mastered incomparably better than any of his 
contemporaries. Yet the historical conclusions he arrives at are 
manifestly absurd on the basis of the data he had himself pre- 
sented. No better proof can be required of his deficiency in 
historical tact. Again, the contrast with Tylor is startling. The 
numerous instances of transmission Tylor adduces, especially 
in the Researches, are not all valid; but none of them is plainly 
absurd, many remain suggestive, if not convincing. 

Turning now to Morgan's specifically sociological work, we 
cannot divorce his interpretations from either his raw material 
or his philosophy of progress. As noted, his picture of society 
was inevitably imperfect from neglect of associational phenom- 
ena, a gap not filled until Heinrich Schurtz's Altersklassen und 
Mdnnerhiinde ( 1902 ) . It was further marred by the biographical 
accident that Morgan's contacts with aborigines began where 
they did — among the democratic clan-organized Iroquois. I have 
often wondered what his scheme might have been like if chance 
had first thrown him among the clanless Paiute, the wealth- 
craving Yurok, the pedigree-mad Polynesians, or the monarchical 
Baganda. Proceeding from the Seneca and encountering for 
hundreds of miles nothing but broadly comparable social struc- 
tures, Morgan prematurely generalized what primitive society 
was like, even though on an apparently wide inductive basis. 
And when he had once formulated the generalization, he could 
dismiss contradictory evidence about the Columbia River tribes 
with the cheap auxiliary hypothesis that their clan organization 
had fallen into decay. I say "cheap" advisedly, because the 
proper procedure — whether definitive or not — on Morgan's own 
principles would have been to examine the kinship systems, 
which he believed always embodied evidence of preexisting 
social conditions. 

On the positive side we note, first, that Morgan personally 
secured much of the North American material, either directly 
or by correspondence; and that his outline of North American 
social organization remained for decades the only comprehensive 
summary. Indeed, imperfect as it is, no adequate substitute in 
the light of present knowledge has yet been provided. Secondly, 
Morgan was not indeed the first to conceive the distinction 


Leiois H. Morgan in Historical Perspective 

between territorial and kinship organization, but he grasped 
Maine's important conceptuahzation ( 1861 ) and made it under- 
he his own treatment of social development. In my judgment, 
Maine and Morgan err in completely denying local ties among 
primitive groups, but doubtless these bonds are comparatively 
weak; and whatever the ultimate verdict may be, Morgan 
deserves credit for recognizing the enormous importance of this 
luminous distinction. 

A genuine contribution was the clearing up of the concept 
"exogamy." Contrary to his wishes, Morgan did not, indeed, 
succeed in banishing the term from scientific nomenclature, but 
his critique of J. F. McLennan in Ancient Society made it pos- 
sible to use it intelligently with reference to observed facts. 
McLennan confused the issue by speaking of tribes as endog- 
amous or exogamous, respectively. Morgan explained that the 
clan, a subdivision of the tribe, was the exogamous unit; and 
that marriage enforced outside this unit normally took place 
within the tribe, so that "both practices exist side by side." 

Another valid stricture advanced against McLennan related 
to his representation of polyandry as a general phenomenon at 
a certain stage, while Morgan rightly regarded it as exceptional 
— a view shared with Lubbock. ^^ 

Various views popularly associated with Morgan need not 
be discussed in detail, because they were not in any sense 
peculiarly his. Thus, the priority of maternal over paternal 
descent had already been postulated by J. J. Bachofen and be- 
come part of the scientific credo of the period. Lubbock, Tylor, 
and Lang were only a few of the scholars who supported this 
view. Again, the notion that individual maniage was inconceiv- 
able among rude savages and must have been achieved by 
gradual progressive stages beginning with sexual communism 
was merely contemporary doctrine. ^^ 

For the same reason I can see nothing remarkable in Morgan's 
ideas on property. According to Professor Matorin, to be sure, 
Morgan's investigations "have proved the communistic char- 
acter of the primitive community and have filled with mortal 
fear the hearts of all obscurants, all those who are guarding the 

'Uhid., p. 151. 
'' Ibid., p. 103. 


Theories and Theorists 

foundations and pillars of traditional morality." ^^ If, however, 
the Russian scholar had looked into Lubbock, he would have 
found much the same position as in Morgan — a firm conviction 
that individual land ownership was always preceded by a period 
in which the land was common." Morgan, of course, attempted 
to fit specific ideas on tliis topic into his general scheme. Thus, 
we learn that savages owned nothing but rude chattels, hence 
had not yet developed a passion for their possession; that at 
first children inherited only from their mother; that slavery — 
a well-established institution on the Canadian West coast — only 
sprang into being in the Upper Status of Barbarism, that is, in 
the Iron Age. At best commonplace, the section of Ancient 
Society devoted to the "Growth of the Idea of Property" is at 
its worst vitiated by the complete neglect of aboriginal slavery, 
aristocracy, and monarchy. 

On the other hand, Morgan, though not ahead of his time in 
this respect, cannot fairly be criticized for ignoring the in- 
numerable instances of incorporeal property that at once nullify 
the theory of primitive communism, since the relevant facts failed 
to arouse interest until a much later date. However for his 
servile followers of today the excuse will not hold. It has, in- 
deed, been asserted that primitive "copyrights" are economically 
insignificant, having merely sentimental value for their holders. 
But this allegation is demonstrably false. A Plains Indian who 
purchases a sacred bundle can readily transmute it into the 
most material of economic goods; he is making the safest of 
investments for the future. Similarly, the Eskimo, communistic 
as they are in the distribution of food, are rabid individualists 
when it comes to magic formulae. "Those who possess the words 
will not part with them, or if they do, it is at a price which 
would soon ruin an expedition." Rasmussen found an old woman 
who had taught a fellow tribesman her spell; in return he pro- 
vided her "with food and clothing for the rest of her life." ^^ 
In modern parlance, she had bought herself an annuity. 

^^ Maturin, "Soviet Ethnography," in Ethnography, Folklore and Archaeology in 
the USSR, IV (1933), 6. 

" Avebury, op. cit., p. 478. 

^® Knud Rasmussen, Intellectual Culture of the Iglulik Eskimos (Copenhagen, 
1929), p. 165. 


Lewis H. Morgan in Historical Perspective 

If Morgan deserves no more credit than a dozen other writers 
for being stimulated by biological evolutionism, he achieved 
unique distinction by his treatment of kinship; his fame rests 
securely and primarily on the Systems of Consanguinity and 
Affinity. In order to visualize the greatness of his achieve- 
ment, w^e must first picture to ourselves the status of relevant 
problems before Morgan. Missionaries and others had, of course, 
noted that primitive peoples classify relatives according to non- 
European norms. But prior to Morgan no one had seen a problem 
in such exotic usage; no one had systematically garnered the 
several nomenclatures, compared them with one another, at- 
tempted a typology or interpretation. Morgan, evincing that 
"eye for essential fact" which even Lubbock was willing to 
concede, devoted twenty years to assembling the pertinent 
facts, partly by personal field studies, partly through the services 
of numerous correspondents. The result is a mass of raw material 
incomparably fuller than anything yet brought together by any 
of his successors. The work represented the terminologies of 
139 distinct tribes or peoples and presented them, on the whole, 
in the greatest detail. Anyone who has worked out a single 
system in the field knows what is implied in this statement. 
Morgan was inevitably limited by the state of ethnographic 
knowledge. Africa was still the Dark Continent, South American 
information was scant or buried in inaccessible sources, from 
which it is only now being disengaged by men like KirchhoflF. 
Having regard to the contemporary situation, Morgan must be 
credited with sparing no pains to achieve a world survey, in 
which he included Indo-European and Semitic as well as native 

The case is illuminating as to Morgan's psychology. Nothing 
is less apt than Mr. Mitra's rhetorical flourish that "like a 
colossus he strode in every field of anthropology." The truth 
is that in most fields he did not stride at all. His contemporaries 
— say, Lubbock — dealt with art, language, and religion; on none 
of these topics can we find any enlightening general ideas in 
Morgan's writings. His interests were narrowly focused, but 
when he dug, he dug deep. And the result is clear. Lubbock's 
versatile mind, outside of prehistory, has left little mark on his 


Theories and Theorists 

successors; while in the one subject to which he devoted him- 
self — social organization and especially kinship terms — Morgan 
remains a towering figure. His work has been revised and am- 
plified, but it cannot be ignored. 

Naturally, the mere assemblage of raw data would never 
confer the title of greatness; nor, may we add, would Morgan 
ever have collected them on so vast a scale unless goaded 
by the search of general principles. It is here, in the appraisal 
of Morgan's interpretations, that discrimination is imperative. 
For, as we have already seen, the assumption of racial affinities 
because of terminological resemblances was not merely wrong, 
but absurd in the light of Morgan's own data. 

Intermediate, however, between ultimate explanation and the 
accumulation of mere data comes the process of classification 
and conceptualization. The qualities Morgan displayed here 
were those characteristic of him in general — not subtletv, but 
. painstaking industry and rugged intelligence. In contrast to most 
y of his successors he saturated himself with the chaotic mass 
of fact, and thus, quite apart from his main purposes, was able 
to foreshadow much of the subsequent typology. He noted the 
aberrant features of the Crow-Hidatsa nomenclature and its 
parallels elsewhere; he personally discovered the "Omaha" tvpe 
of system among the Kansas and indicated its distribution; im- 
perfectly informed as he was about our Western tribes, he 
detected such anomalies (from the Iroquois angle) as the dis- 
tinction of paternal and maternal grandparents and the use of 
reciprocal terms. ^^ Incidentally Morgan thus provided a solid 
basis for distributional studies and discoveries of historical con- 
nection, even though of an order less pretentious than that he 
consciously envisaged. 

In the conceptualization of types of relationship terminology 
Morgan was not wholly successful. He was hampered by defec- 
tive knowledge of those very tribes whose systems had a crucial 
bearing on the problem of classification. The data from Eastern 
and Central North American tribes, eked out by Oceanian and 
Australian reports, inevitably encouraged the inference that aU 
aboriginal tribes grouped relatives into large classes. Hence, 

"Morgan, Systems, pp. 179, 188, 211, 245, 247, 252 f., 262. 


Lewis H. Morgan in Historical Perspective 

Morgan's main category of Classificatory Systems, to which he 
opposed that of Descriptive Systems as supposedly characteristic 
of the Indo-European family. Rivers pointed out that "descrip- 
tive" was not an apt term since most English terms, for in- 
stance, are denotative rather than descriptive. To be sure, 
Morgan's contention was rather that the original "Aryan" system 
was descriptive; that, like Erse, it had denoted a few primary 
relationships by specific nouns and described all others, that is, 
"son of brother," "son of son of brother." But obviously even on 
that assumption Rivers was right in objecting that modem 
Indo-European languages for the most part failed to fall into 
Morgan's rubric. From the opposite side Kroeber has urged that 
English and related tongues are not devoid of classificatory 
terms, such as "cousin." To me, however, there seems to be a 
still more vital objection. "Descriptive" designates a technique 
for defining relationship, "classificatory" a mode of grouping. 
The two concepts are thus not mutually complementary but 
relate to different subjects of discourse. Hence, a descriptive 
phrase, say, father's brother's son, might quite conceivably be 
applied to an indefinite number of individuals. 

At present we are, indeed, still far from a satisfactory group- 
ing of nomenclatures, but certain conclusions seem definitive. 
A nomenclature of relationship is not usually one system, but the 
result of several crisscross currents: as ICroeber phrases it, there 
are a number of diverse categories. ^'^ Morgan fully realized the 
significance of one category, that of grouping collateral with 
lineal kindred, but he rather naively assumed that features 
wliich merely chanced to appear with this fundamental trait 
were organically bound with it. Thus, the separation of junior 
from senior siblings constantly figures as a criterion of "clas- 
sificatory" systems. Yet obviously the aborigines in question are 
not thereby making classes larger than ours, but on the con- 
trary are introducing distinctions foreign to us. If the temi has 
any significance at all in such a context, it is we who are clas- 
sificatory; and, however that be, the segregation of senior from 
junior siblings is not possibly related — except by historic ac- 

^ A. L. Kroeber, "Classificatory Systems of Relationsliip," RAI, Journal, XXXIX 
(1909), 81. 


Theories and Theorists 

cident — to the inclusion of parallel cousins with sibHngs or of 
paternal uncles with the father. 

Morgan's strength and weakness appear clearly in his treat- 
ment of the Eskimo "system." With exemplary candor he points 
out that eight of the ten features indicative of the North Ameri- 
can Indian terminologies are lacking. Nevertheless, he defines 
it as classificatory. "It is ... a classificatory as distinguished 
from a descriptive system. But in the greater and most important 
fundamental characteristic of this system it is wanting. The 
Eskimo form not only fails in the necessary requisites for the 
admission of this people, upon the basis of their system of 
relationship, into the Ganowanian family, but furnishes positive 
elements to justify their exclusion." There may possibly be, he 
argues, an ultimate, there is certainly not an immediate, relation 
of the systems; indeed, the Eskimo nomenclature approaches the 
Aryan and Uralian more closely than it does that of the American 

Comparing this statement with Frazer's glib reference to the 
Eskimo terminology as "classificatory," we at once recognize 
Morgan's immeasurable superiority. There is all the difference 
in the world between a bowing acquaintance and a serious 
grappling with refractory kinship data. Morgan has not achieved 
conceptual clarity, but he knows that here is something different 
from the Seneca norm, something so different that it verges 
on the exact opposite of a classificatory system, even if he 
cannot quite muster up courage to call it "descriptive." 

Apart from his blind spot as to racial aflBnities, Morgan by 
concentration on his schedules was preserved from errors to 
which more casual investigators have fallen prey. His rebuttal 
of McLennan, for example, refutes the quaint notion that kin- 
ship through females implies ignoring of patrilineal kin, a con- 
clusion apparently not yet universally accepted. 

The distinction of Morgan, then, is not simply that he heaped 
up vast stores of information on a subject of theoretical import, 
but that he immersed himself in this welter of fact, came to 
grips with it, thought about it. The specific quality of his think- 
ing shows to advantage in the discussion of the clan organiza- 

^"^ Morgan, Systems, pp. 277, 470, 510. 


Lewis H. Morgan in Historical Perspective 

tion as a possible cause of the Seneca type of classificatory 
system. He examines the' imphcations of such an organization 
and conclusively demonstrates that, while it would account for 
the identification of sisters' children with siblings under maternal 
descent, the similar identification of brothers' children would 
remain unexplained.^^ So, of course, it does except in the 
specialized case of exogamous moieties. 

Of Morgan's scheme for the evolution of the family, little 
need be said, because, as already suggested, it contained little 
that distinguished him from the other evolutionists. Promiscuity 
naturally came first, but avowedly not as anything but a the- 
oretical deduction; the monogamous family came last; and 
there were intermediate stages bridging the gap. Morgan's orig- 
inality lay in bringing this scheme of stages into correlation with 
forms of kinship terminology. That he committed a number of 
grave errors in this connection is pretty generally admitted. In 
harmony with the spirit of the times he took it for granted that 
what was simpler must be older, hence inferred that the Poly- 
nesian nomenclatures are more archaic than their American 
equivalents. He made a really fatal error in supposing that 
when a Polynesian addressed his father and, say, his mother's 
brother by a single term this implied conceiving the uncle as 
a possible procreator. Objectively formulated, the fact simply 
is that one common term applies to the begetter and other male 
relatives, that the Polynesians lack a term denoting paternity 
in our sense. Hence, Morgan's conclusion that the maternal uncle 
once mated with his sister (the speaker's mother) is fallacious: 
that custom would indeed logically produce the observed clas- 
sification, but it is not the only possible usage that can lead to 
this result. Members of the same sex and generation may simply 
be grouped together under a blanket status term. Morgan's 
mistake, then, lies in misinterpreting the import of the ab- 
original facts by reading a modern meaning into tlie translated 
kinship terms and in ignoring alternative detenninants. 

When, however, we discount a pioneer's pitfalls and the warp- 
ing due to contemporary bias, a magnificent and valid conception 
remains. Lists of relationship terms are lexical elements devoid 

''Ibid., p. 476. 


Theories and Theorists 

of interest except to a linguist unless they are brought into 
contact with the elements of reality to which they refer. What 
is more, the linguistic approach culminates not in an explana- 
tion of the phenomena, but in a negation of the possibility of 
explanation. Kinship nomenclatures are certainly amenable to 
the changes which affect words, and since these changes are 
capricious we can never hope to reduce all features of relation- 
ship terminology to social antecedents. However, insofar as they 
are explicable at all they must be explained on sociological lines. 
The Southern Siouans and the Miwok of central California 
belong to diverse linguistic families, are separated by a distance 
of well over a thousand miles, are not known ever to have 
lived in close proximity. Their nomenclatures share the "Omaha" 
features of classing a mother's brother's son with the maternal 
uncle — a feature lacking in all the intervening tribes. Do the 
Southern Siouans resemble the Miwok more than their fellow 
Siouans of Montana and Dakota because of a miracle, or is it 
because both Southern Siouans and Miwok demonstrably share 
forms of marriage and linked social customs not found among 
the Crow and Teton? The choice lies between a sociological in- 
terpretation and the abandonment of interpretation. 

We thus find Morgan's major postulate vindicated that kin- 
ship terminologies in some measure correspond to social facts, 
among which matrimonial rules are prominent. However, Morgan 
also asserted that social custom advanced while its lexical 
equivalents remained stationary; the use of a kinship term 
thus may point to the prior existence of an obsolete usage. 
Some who accept the correlation of social custom \^ith nomen- 
clature balk at this application of the principle of survivals. 
y Trained to view "survival" arguments with suspicion, I have 
become convinced that the avowed skepticism on this point 
harbors as much cant as the evolutionary zeal of our predeces- 
sors. Indubitably cultural changes proceed with uneven velocity, 
hence certain elements lag while others spurt ahead; further, 
linguistic phenomena are markedly conservative. These accepted 
facts warrant the assumption that a terminological feature in 
harmony with a certain custom may survive that custom. The 
only question is whether the social factor is the only possible 


Lewis H. Morgan in Historical Perspective 

determinant, whether the really vital factor is not rather one of 
its correlates, whether the same result may not be effected by 
a different cause. But when due allowance is made for this, 
Morgan's principle of survivals remains a valuable procedure. 

In accepting the independent repetition of terminological 
features where social concomitants are repeated, we avowedly 
admit "evolution." However, it is a strictly empirical parallelism, 
which does not pretend to sketch culture history as a whole, 
but merely to account for specific resemblances. It is, indeed, 
very difficult to deal with cultural phenomena and fail to 
recognize certain organic bonds between phenomena. As Father 
Sclimidt has recently explained, aprioristic evolutionism must 
be eschewed, but it is quite proper to make a "quite logical 
deduction from the very nature of things and men, to arrange 
them [cultural measures] in a certain series of phases of develop- 
ment." ^^ Morgan certainly believed in the logical character of 
his deductions, and we therefore criticise him mainly for his 
deficient knowledge of the "nature of things and men." 

Ultrascientific critics like to remind us that cultural phenomena 
are very complex, that we are consequently not warranted 
in assuming more than a functional relationship between distinct 
traits. There is great merit in the contention, but like every- 
thing else it can be overdone. If South African natives and 
North American Indians respond to the impact of white civiliza- 
tion by quite similar Messianic cults, shall we not admit that 
the contact is a temporally antecedent cause? Or do the meth- 
odological wiseacres contend that the cults could bring about 
the invasion of the white race? 

Morgan thus figures as a typical exponent of a contemporary 
philosophy of civilization. That philosophy was bedecked with 
the follies of fashion, but part of its core was sound and a dis- 
criminating analysis will try to preserve it. As an individual, 
Morgan was only moderately cultivated and indifferently familiar 
with culture history. Most emphatically, he was not, as he 
has been called, a man of "brilliant delusions." He had delusions, 
which he set forth with schoolmasterly pedantry; but he bril- 

^ W^ilhelm Schmidt, "The Position of Women with Regard to Property in Primi- 
tive Society," American Anthropologist, XXXVII (1935), 244-256. 


Theories and Theorists 

liantly illuminated the subject of kinship terminologies by im- 
mersing himself in the facts, persistently arranging them, seek- 
ing and rejecting specific solutions of the problems they pre- 
sented. His was not a flashy intellect, but one of unusual honesty, 
depth, and tenacity; and his prolonged concentration achieved 
the triumph of glimpses of real insight in a virgin field of scholar- 
ship. There is no better example of Darwin's saying, "It's dogged 
does it." 



Cultural Anthropology: A Science 

Every well-defined segment of reality calls for a 
branch of knowledge that shall record and interpret its phe- 
nomena. During the last hundred years it has become increas- 
ingly clear that culture — the sum total of socially acquired 
thoughts and practices — represents such a distinct domain; and 
the discipline dealing with it has been variously ticketed "culture 
history," "ethnography," "ethnology," or "cultural anthropology." 
This branch of learning does not confine itself to "primitive" 
tribes because the lines of cleavage between them and civilized 
peoples are arbitrary; also because a social tradition is equally, 
distinctive of modern America and the Chinook Indians, of 
ancient Athens and the Australian aborigines. Whatever broader 
principles may emerge from a study of culture should therefore 
hold in some measure for all its levels. 

The same subject matter can, of course, be approached from 
different and equally legitimate angles. Horses may be painted 
by a Rosa Bonheur or dissected by a comparative anatomist. 
So a culture may be described by a literary artist or by a 

Read at the annual dinner of the Sigma Xi and Phi Beta Kappa societies of the 
University of Oregon, May 23, 1936, and printed in American Journal of Sociology, 
XLII (1936), 301-320. 


Theories and Theorists 

scientist; their interests do not clash, because except by chance 
they do not meet. The artist's task is to convey an aesthetic 
impression, hence he rightly omits whatever might mar that 
effect; and he has no call to explain why the culture is what it 
is. The investigator, on principle, omits nothing because he ought 
to register the whole of his reality; and he must coordinate his 
data in spatial, temporal, and functional terms. 

There is a widespread but, in my opinion, vicious tendency 
to dissociate social from natural science. On the one side we 
hear that the social sciences must consider values, hence sub- 
jective elements that militate against an objective approach. 
Again, their data are said to be so complex as to preclude 
generalizations. Some also stress the impossibility of experiment, 
the absence of mathematics. 

Each of these objections rests on a misconception. Ethnology, 
for instance, must indeed deal with values, for they are part of 
its subject matter and to ignore them would be to neglect part 
of cultural reality. Thus, in Australia, as in many areas of Africa 
and America, a woman and her daughter's husband are forbidden 
to converse or even look at each other. The associated senti- 
ments are part and parcel of the phenomenon. Translating into 
modern terms, we might easily view it in terms of our mother- 
in-law jokes. Nothing, however, would be farther from the 
truth, for wherever the aboriginal attitude has been determined 
it turns out to be one of mutual respect expressing itself in 
avoidance. This insight into the emotional states involved is, 
however, not the result of a magical empathy but rests on the 
utterances of native witnesses, which can be checked by further 
inquiry. Metaphysically it is of course conceivable that in- 
formants will lie about their sentiments; but, even so, if they 
all told the same story, such uniform prevarication would itself 
be a fact of importance. At all events, there is no other way of 
arriving at the truth in the matter than by questioning the 
natives and noting their behavior. Science determines subjective 
phenomena by objective methods. 

Fortunately, human beings often voice their sentiments spon- 
taneously, and in the collections of primitive prose now available 
there is abundant evidence as to the nature of personal relation- 


Cultural Anthropology: A Science 

sliips. One tale, for example, brings home the sacredness of 
comradeship when a father disowns his son for infidelity to 
his friend/ The traditions of several tribes incidentally demon- 
strate what has so often been denied — the occurrence of romantic 
love on the primitive plane. A Plains Indian lover is separated 
from his sweetheart. "As he sat there, he felt as though he must 
die of grief." The girl "also was very heavy of heart." She falls 
ill, and though the young man reaches her bedside she dies, 
leaving him as a souvenir a pair of beautiful moccasins em- 
broidered by herself. "And the young man, caring nothing for 
the others who sat in the tipi looking at him, broke down and 
wept." When the other people have moved on, he enters the 
tipi erected over her burial scaffold, and as a reward for his 
devotion she is allowed to revive and to marry him.^ The theme 
of Orpheus and Eurydice, widespread in North America, gives 
similar evidence. To quote a Menomini version: "A certain man 
was married to a woman; greatly they loved each other. Tf you 
die first, I shall go with you,' he would say to that wife of his. 
And she, too, then would say the same to her husband." The 
wife actually dies, and her husband promptly follows her to the 
hereafter. So in a central Californian variant the mourning hus- 
band goes to the grave and declares, "I'm going to stay here. 
I'm going to watch you. Where you're going, I'm going." ^ 

Besides tales there are other sources for the true inwardness 
of native life. The autobiographies secured by Drs. Paul Radin, 
Julian H. Steward, and Truman Michelson contain priceless 
sidelights on native feeling and thought.^ The important point 
about both these personal narratives and the fictitious folk tales 

^ R. H. Lowie, The Assiniboine, AMNH, Anthropological Papers, IV (1909), 

^Ella Deloria, Dakota Texts (New York, 1932), pp. 224-232. 

=* Leonard Bloomfield, Menomini Texts (New York, 1928), p. 125; cf. A. H. 
Gayton, "The Orpheus Mydi in North America," Journal of American Folk-Lore, 
XLVIII (1935), 263-293. 

* Raul Radin, Crashing Thunder ( New York, 1926 ) ; Truman Michelson, "Auto- 
biography of a Fox Indian Woman," BAE, 40th Ann. Rept. (1925), pp. 295-349; 
idem, "Narrative of an Arapaho Woman," American Anthropologist, XXXV ( 1933), 
595-610; idem, "The Narrative of a Southern Cheyenne Woman," Smithsonian 
Miscellaneous Collections, LXXXVII, No. 5 (1932), 1-13; J. H. Steward, Ttvo 
Paiute Autobiographies, Univ. Calif. Publ. Amer. Arch, and Etlui., XXXIII ( 1934), 


Theories and Theorists 

is that they yield objective evidence about subjective states. 
For what we want to ascertain is not how ethnologist X responds 
to the Winnebago scene, but what the Winnebago themselves 
feel; and with reference to their intimate life, that can be satis- 
factorily established only by recording their very words. 

To prevent misunderstanding, I add that every addition to 
our stock of knowledge about aboriginal attitudes should be 
heartily welcomed. Thus, it is one of Boas' outstanding achieve- 
ments to have directed attention to the primitive artist's psycho- 
logical position toward his art. Not content to study museum 
specimens, Boas injected the question of how the innovator 
works on the basis of his tradition and how his environment 
receives his creations.^ But here once more the point is to 
obtain the ipsissima verba of the creators and of their critics. 
To depart from this ideal is to glide down the inclined plane 
of rank impressionism. 

Other supposed reasons for an antithesis of anthropology and 
science rest on a misapprehension of natural science. Laymen 
readily hoodwink themselves into believing what no serious 
natural scientist claims, viz., that all natural science is con- 
cerned with Newtonian universals. Patently false for geology 
and the biological sciences, this proposition is not even true of 
physics, chemistry, and astronomy. Astronomers can predict 
solar eclipses, but they candidly admit their inability to predict 
meteoric showers. And the physicist, regardless of his lay touts, 
is usually satisfied with generalizations of moderate scope. When 
he is confronted with an irreducible fact, such as the expansion 
of water in freezing, he accepts it; he does not argue that it 
cannot be true, nor does he insist that it is worthless because 
other liquids behave differently. The position of the scientific 
anthropologist is precisely the physicist's and the astronomer's: 
he determines his data with the greatest precision feasible, 
generalizing about them exactly so far as the nature of the 
case permits and no further. As for experiments, the crucial 
experiment is not that in the laboratory but the Gedankenex- 

^ Franz Boas, Primitive Art (Oslo, 1927), pp. 155 ff.; Franz Boas, H. K. Haeber- 
lin, J. A. Teit, and Helen H. Roberts, "Coiled Basketry in British Columbia and 
Surrounding Region," BAE, 41st Ann. Kept. (1928), pp. 119-484. 


Cultural Anthropology: A Science 

periment, and that is quite as practicable in the field of culture 
as anywhere else. In short, cultural anthropology is simply 
science grappling objectively with one aspect of the universe and 
inevitably limited in its determinations only by the nature of 
its data — no more uniformly refractory than the astronomer's — 
and the provisional inadequacy of its present techniques. Its 
being in part historical does not make it antithetical to natural 
science, for geology, biology, and even astronomy are largely 
devoted to the study of sequences. The failure to employ math- 
ematics on a major scale draws no line of cleavage, for here 
again there are natural sciences that use the calculus with 
equal restraint. Of organic chemistry. Professor Lewis declares: 
"The whole theory of structure requires about as much mathe- 
matics as a child needs for building houses with blocks." ^ 

The first task of ethnology is to determine the spatial rela- 
tions of all phenomena that are passed on from generation to 
generation by social inheritance. Nowadays it is fashionable 
to deride such investigations on the plea that they wrest an 
institution or industry from its natural setting. Each culture, 
it is averred, is a closed system to be viewed only as an integral 

As a matter of fact, this position was at one time opportune, 
when antiquarians were gleaning oddities of custom with the 
fervor of a philatelist collecting stamps. As a counterblast the 
totalitarian point of view was unquestionably helpful and timely, 
as proved by its many champions of otherwise diverse persuasion 
— men like Professors Boas, Radcliffe-Brown, Thurnwald, Mali- 
nowski, Durklieim. This relatively great service should not blind 
us to the doctrinaire nature of the propositions urged. 

In the first place, a specific culture is an abstraction, an ar- 
bitrarily selected fragment. Social tradition in some measure 
varies from family to family. Shall we, then, study a single Hopi 
family, the Hopi village of Walpi or the seven Hopi villages, 
the Pueblo area of North America, if not the New World as 
contrasted with the Old? There is only one cultural realitv that 
is not artificial, to wit: the culture of all humanity at all periods 
and in all places. The choice of a particular people is accord- 

* Gilbert N. Lewis, The Anatomy of Science (New Haven, 1926), p. 172. 


Theories and Theorists 

ingly a limitation justifiable by expediency only; and there is 
no warrant for the a priori assumption that facts pertaining 
to Walpi are mystically connected only with other Walpi facts 
and with nothing else in the universe. 

To be sure, bonds between intra-tribal elements have been 
established, for example, in the way of ceremonial patterns, but 
these have never been proved to cover more than a restricted 
portion of the empirically cohering features. Thus, the Arapaho, 
a Wyoming tribe, perform their several ceremonies only after 
a formal vow. This suggestive fact helps us understand why any 
new ceremony, whether borrowed or evolved within the tribe, 
would conform to the established standard. If this phenomenon, 
however, is somehow related to the taboo among the Arapaho 
against free chatting between brother and sister, the lack of 
clans, the smoking of catlinite pipes, and the use of snowshoes 
in buflFalo-hunting, the correlation has never been demonstrated, 
and in my opinion it never will be. 

The gist of the matter is simply this. Isolated facts yield no 
insight; but nothing warrants the assumption that all significant 
relationships are confined within the system of a particular 
social tradition. To wrest a fact from its tribal context is no more 
arbitrary than to wrest the tribe from its contacts with the rest 
of humanity; than to isolate the extremities of vertebrates for 
comparative anatomical study when we all know that the limbs 
never exist except in association with the organism as a whole. 
A commonplace consideration may clinch the argument. The 
conception of the zero, as well as the arithmetical notation 
linked with it, has rightly been acclaimed as a capital achieve- 
ment of the Maya mind. But what lends peculiar importance 
to the phenomenon? Not that the idea is imbedded in a cere- 
monial calendric system; irrational contexts of rational notions, 
once interesting, have become trite. But that the Indians of 
Yucatan were able to excel the Greeks and Romans — that is, 
indeed, a datum of primary significance, regardless of anything 
else we know about the Maya, and bearing in the most definitive 
manner on the racial potentialities of the Indian. 

Years ago I wrote: "When we know only the range of a usage, 
we may not yet know very much, but we have at least a point 


Cultural Anthropology: A Science 

of departure for amplifying our inforaiation. When we do not 
know the distribution of a phenomenon . . . , we know nothing 
that is theoretically significant." The statement was bitterly 
resented in some quarters, but the Maya case illuminates the 
intended meaning. Assume that the Australians and Andaman 
Islanders shared the position system with the Maya; obviously 
the fact would be seen in quite a di£Ferent light. Or assume 
that the trait occurred generally among farming peoples but 
never in the pastoral or hunting condition; the import of the 
datum would again be altered. 

The superb distribution studies of the late Baron Norden- 
skiold and of Professor Leslie Spier are invaluable contribu- 
tions to a science of culture not because knowledge of the 
geographical extent of traits is an end in itself but because 
accurate information on distributions raises problems of a basic 
character and helps toward their solution. Geographical con- 
siderations merge in those of a temporal and causal order. We 
may or we may not find an elusive all-integrating intra-tribal 
factor, but we must correlate significantly within and without 
the culture, and for that the distributions must be known. Why, 
for instance, is skin-dressing a masculine pursuit among our 
southwestern natives when elsewhere north of Mexico it is 
regularly a task for women? Spier notes that Pueblo men weave 
and plausibly suggests that skin preparation was accordingly 
transferred to them.'^ Again, Nordenskiold remarks that in South 
America the carrying-net or shawl occurs in the Inca empire 
while the Amazonian tribes use carrying-baskets. In the former, 
flour is prepared by grinding on stone hand-mills; the other 
substitutes pounding in a wooden mortar, etc.^ We know that 
many features have been diffused from Peru to adjacent ter- 
ritories; why have the Amazonians remained immune in the 
instances cited? Similarly, in the United States two forms of 
guessing-game, the moccasin and the hand-game, are in the 
main mutually exclusive — somewhat as in South Germany tarok 

^Leslie Spier, Havasupai Ethnography, AMNH, Anthropological Papers, XXIX 
(1928), 157. 

*Erland Nordenskiold, Comparative Ethnographical Studies (Goteborg, 1930), 
VII, 134. 


Theories and Theorists 

cards hold their own against the North German skat. Here is 
one of the ethnologist's perennial puzzles — the process of selec- 
tive borrowing. What underlies it? Why will people take over 
only part of the alien traits to which they are exposed? Nor- 
denskiold gives a sound partial explanation: People do not 
supplant a well-tried device fulfilling a particular need. The 
distributional data, at all events, lead us into the very heart 
of our problems; imperceptibly we gUde from "where?" to 

Distributional phenomena lead to a study of man's adapta- 
tion to his physical environment. The extravagant claims of 
certain geographers — by no means all of them — have unduly 
prejudiced many ethnographers against the geographical point 
of view. Nothing is easier than to refute the notion that man 
automatically responds to a given practical problem in the 
most elegant fashion. In New Mexico and Arizona sedentary 
farmers living in stone houses have lived for centuries beside 
sparingly horticultural nomads building crude huts; and in the 
Gran Chaco the simpler Choroti inhabit the same territory as 
the Chiriguano, who are deeply tinctured with Inca influences. 
The Ona of Tierra del Fuego freeze in an Antarctic habitat 
because they have not contrived adequate dress or shelter, etc. 
Just as soon, however, as we cease to attach messianic notions 
to geography and treat it in a common-sense way, it becomes 
a helpful and necessary guide; and significant correlations emerge, 
partly on the intra-cultural plane, partly with outside factors. 

The negative influence of environment has been generally 
recognized, but many overlook the positive complement of 
such inhibition. In the Chaco and on the upper Xingu the 
dearth of suitable material precludes a Stone Age. The ab- 
origines, requiring some tools, substituted shell, bone, teeth, 
and hardwood, just as the Micronesians in a similar predica- 
ment made adze blades of the giant clam shell. What is more, 
geographical limitations are among the major motives for trade; 
and with commercial relations once established for a specific 
object, a contact is created that may lead to many other gains. 
Australians who go in search for diorite or the pitjuri stimulant 
also bring home songs, dances, and tales. In tropical South 


Cultural Anthropology: A Science 

America the curare poison became an important article of trade. 
This led to a change of weapons in some regions where the 
blowgun would be ineffective in warfare and the hunting of 
large game; tipped with strychnine, its darts turned into deadly 
projectiles.^ Environment is thus no mean determinant of cul- 
ture. We must beware of an intolerantly all-or-nothing position. 
Though geography is impotent to explain many phenomena 
we should like to understand, it is nevertheless capable of partly 
accounting for others, and for data admittedly complex we 
have no right to demand a single simple cause. 

The distributional facts thus prompt basic problems as to 
why and how cultures differ, and also how one element in a 
geographico-ethnic situation may affect others. They further 
suggest questions of a historical order, i.e., of sequence and of 
tribal relations. Here we encounter a major objection, since 
some writers reject such matters as trivial. For scientific anthro- 
pology the rejoinder is simple. Since it embraces all the phe- 
nomena of culture in their varied relations, temporal succession 
is manifestly important because time is one of the inescapable 
categories of our thinking. We cannot picture a timeless culture 
any more than a spaceless one. The very occurrence of farming 
in contemporary France dates its culture, for in a Reindeer age 
it would have been impossible. Migration either of peoples or 
of technical processes is likewise inconceivable in certain areas 
during the boatless period of mankind unless there were land 
bridges, which again involves a geological time scale. As for 
tribal relations during the geologically Recent period, they are 
significant in several ways. Whether a parallel in two distant 
cultures is due to independent development or must have sprung 
from one source has a direct bearing on the nature of human 
or racial inventiveness — a question of manifest importance for 
an ultimate theory of culture. Further, since diffusion is an ac- 
cepted fact, judgments varying merely as to its extent, we must 
amass all information we can about its conditions. From this 
angle the mere fact of dissemination is only a starting-point. 
We ought to ascertain who borrows from whom, noting the 

° Karl von den Steinen, Unter den Naturvolkern Zentral-Brasilicns ( Berlin, 
1897), pp. 196 ff.; Nordenskiold, op. cit., I (1919), 4, 15; III (1924), 4, 59 f. 


Theories and Theorists 

comparative planes of the recipient and the donor; as aheady 
suggested, we must inquire why part of the possible features 
is rejected, and how the accepted offerings are assimilated. 
These are questions more easily asked than answered, yet an 
adequate theory cannot ignore them. 

In the meantime we may record some gratifyingly sohd 
results of historical studies. Archaeology, like paleontology, labors 
under certain limitations, but, like its sister-science, yields defi- 
nite results. Metal ages do not occur everywhere in the same 
order, but everywhere metallurgy was preceded by work in 
stone, bone, wood, or shell. Hunting and gathering uniformly 
antedate farming. To be more specific, the ancient Peruvian 
coastal site of Paracas harbors only one variety of maize char- 
acterized by small dark-brown or black cobs; manioc and sweet- 
potato tubers found there are small and twisted, being similarly 
inferior to later samples. ^° Here, as in the more familiar case of 
our southwestern Indians, we can thus follow the evolution of 
agriculture. Stratigraphic investigation has also repeatedly traced 
the succession of stone implements, of types of container, of 
pottery forms and pottery ornamentations. Even surface finds 
may suffice for establishing a sequence. In modern Amazonas, 
stone arrowheads are completely lacking, yet they are not rare 
in old settlements. The Indians evidently had stone implements 
when they entered tliis region, were unable to perpetuate stone 
work for want of raw material, and substituted bamboo or hard- 
wood. Similarly, the Chaco tribes manufacture bone or wooden 
knives simulating the form of stone blades unearthed in ar- 
chaeological sites farther west.-^^ 

Naturally, sequences established without the aid of archae- 
ology are generally less definitive, yet when worked out with 
due caution they may be accepted as contributions to knowl- 
edge. In any event they are not one whit more dubious than 
the comparable findings of natural science. Some of us are old 
enough to remember Lord Kelvin maintaining the physicist's 
point of view that the age of the earth was insufficient to allow 

"" Nordenskiold, op. cit., IX (1931), 51. 

"^Ihid., VIII (1930), 33 f.; cf. Sigvald Liime, Darien in the Past (Goteborg, 
1929), pp. 54 f. 


Cultural Anthropology: A Science 

for biological evolution. If modern estimates are right, the great 
British scientist was incomparably farther from the truth than 
the wildest speculators about culture history. 

Nordenskiold has drawn attention to an interesting distribu- 
tional phenomenon. Comparing the two parts of the New World, 
he finds a large number of North American features recurring 
in South America, but only in the Chaco and regions south. 
More recently Krickeberg, without restricting himself to the tip 
of South America, has added appreciably to the parallels. Some 
of these, he believes, may have simply spread from tribe to tribe 
at various periods; others he ascribes, like Nordenskiold, to a 
very ancient American layer, seeing that they are not found in 
the intervening areas. ^^ This argument is indeed definitive, on 
two conditions. First, the features must not be lacking from 
mere failure on the part of the observers; second, they must be 
so well defined as to bar independent development. Neither 
Nordenskiold nor Krickeberg seems to me to have sufficiently 
observed these precautions; but when due allowance is made, 
some striking parallels remain. Thus, a communal game drive 
by firing the grass is reported for the Algonkian and Siouan 
tribes of Wisconsin, Illinois, Iowa, North Dakota, and Min- 
nesota; for some eastern and southeastern Indians of the United 
States; for the Havasupai of Arizona, the Coeur d'Alene of Idaho, 
and several North Californian tribes. ^^ According to Krickeberg, 
this practice was shared by the Brazilian Ge, the Abipone and 
Pilaga of the Chaco, and the Patagonians. This trait I should 
concur in assigning to a very old cultural layer, possibly a 
proto-American one, for we know that communal hunts are 
ancient and in such battues fire would be a most effective agent 
in driving large game toward a cliff or pound. Once invented, 
the method would not be readily abandoned. Obviously, how- 
ever, the tropical forest does not provide a favorable terrain, 
nor would preponderantly horticultural peoples tend to per- 

" Nordenskiold, op. cit, VIII (1930), 127 f.; IX (1931), 77-94; W. Krickbercj, 
"Beitrage zur Frage der alten kultiirgeschichtlichen Beziehungen z\vischen Nord- 
und Siidamerika," Zeitschrift fur Ethnologie, LXVl (1934), 287-373. 

" Clark Wissler, Material Culture of the BJackfnot Indians, AMNH, Anthro- 
pological Papers, V (1910), 51; Kaj Birket-Smith, The Caribou Eskimos (Copen- 
hagen, 1929), II, 330 f. 


Theories and Theorists 

petuate such drives. Thus, both the positive and the negative 
instances are satisfactorily subsumed under the hypothesis that 
the New World drives with grass-firing represent one historical 
phenomenon belonging to a very early hunting culture. 

For corresponding reasons I share Nordenskiold's view that 
the far-flung North American mode of boiling by dropping hot 
stones into a container of basketry, wood, hide, or bark passed 
into South America, surviving only among the Chono of Chiloe 
Island and the Kaingang of southeastern Brazil, with some sug- 
gestions of the practice in Tierra del Fuego.^^ 

For a number of chronological determinations anthropology 
relies on zoological and botanical results. Thus, with some ex- 
ceptions the American aborigines from Alaska to Tierra del 
Fuego have do2;s. Have these canines been independently do- 
mesticated in the several regions from local forms of wolves, 
covotes, or foxes? Or do they represent an import from the 
Old World? Comparative anatomists assure us that aboriginal 
dogs cannot be descended from American wild forms, but one 
and all go back to a single Asiatic wolf.^^ By itself this does 
not exclude comparatively recent transmission from Asia, but 
such an assumption is impossible. In the first place, canine 
remains occur in archaeological sites of obviously considerable 
age. Second, notwithstanding their common origin, the pre- 
Columbian dogs of the continent represent some sixteen distinct 
breeds, differentiation indicating long domestication. Again, the 
beast occurs among the very simplest tribes and those most 
remote from the point of entry — such as the Basin Shoshoneans 
and the Fuegians. 

Utilization of the species shows great variabilitv. Dog traction 
is found with sledges among the Eskimo, with the travois or 
frame connecting dragging poles on the Plains, but is lacking 
farther south. The Dakota and Arapaho ate dogs, a custom 
positively abhorrent to the Yahi of northern California, who 

" Nordenskiold, op. cit., IX (1931), 86; Hermann Ploetz and A. Metraux, "La 
Civilisation materielle et la vie sociale at relisjieuse des indiens Ze du Bresil 
meridional et oriental," Revista del Instituto de Etnologia de la Universidad 
Nacional de Tucumdn, I (1930), 170 f. 

^ Glover M. Allen, "Dogs of the American Aborigines," Btdletin of the Museum 
of Comparative Zoology, Harvard, LXIII (1920), 431-517. 


Cultural Anthropology: A Science 

considered the flesh deadly poison craved by evil sorcerers for 
their diabolical practices. These people hunted bear and deer 
v^ith dogs; and to the Ona of Tierra del Fuego these animals 
are a virtual necessity for bringing down guanacos. In north- 
eastern Bolivia, on the other hand, dogs are of no practical im- 
portance. Emotional attitudes are equally variable. The Pausema 
of Brazil starve their pariah dogs, while a Choroti woman in 
the Chaco will suckle a pup beside her own infant. In Arizona, 
again, the Maricopa treat dogs as persons, give them names, 
and dream about them. 

Such biological and cultural differentiation then, supported 
by archaeological data, indicates that the dog belongs to a very 
ancient American stratum. Introduced by early immigrants from 
Asia, it gradually spread southward, developed local breeds, 
assumed varying uses in different regions, and stimulated dis- 
tinctive attitudes in diflferent areas. 

The pan-American approach need not be limited to material 
culture. In comparing myths some investigators treat vague 
similarities as significant, but that is a peril of all scientific com- 
parison. Obviously we must also guard against mistaking a post- 
Columbian parallel due to common Negro or European influences 
for resemblance proving ancient contact. A tale that has im- 
pressed me as really pertinent to the argument embodies the 
dualistic conception of mythical heroes, generally brothers, some- 
times twins, one of whom endeavors to create for mankind a 
life eternal of ease and plenty but is thwarted by his marplot 
r^omrade. Gusinde has suggestively discussed the South Ameri- 
can myths of this category, but without pointing out what is to 
me the most remarkable phenomenon, viz., the difl^erential re- 
semblances of the Yaghan and Ona versions with those from 
California and the Great Basin. Since these last two regions are 
geographically contiguous and demonstrably have close cultural 
relations, they form a unit as against the Fuegians.^*^ 

The resemblances are of a specific character — a conclusion 

" Martin Gusinde, "Das Briiderpaar in der sUdamerikanischen Mythologie," 
XXIII International Congress of Americanists (Lancaster, Pa., 1930), pp. 687- 
698; Wilhelm Koppers, Unter Feuerland-lndiancrn (Stuttgart, 1924), pp. 202- 
208; Alfred Metraux, La Religion des Tupinamba (Paris, 1928), pp. 31-43. 


Theories and Theorists 

forcibly brought home by the much wider divergence between 
the Yaghan and the Tupi-Guarani twin stories than between the 
Yaghan and the Cahfornian-Basin versions. Obviously it is not 
a matter of psychic or of pan-American unity. There is a dif- 
ferential resemblance, such as confronts the naturalist who finds, 
say, a marsupial in Australia and again in America. The eth- 
nographer, of course, does not infer a direct genetic kinship be- 
tween Fuegians and Californians but merely a pristine contact 
producing an exchange of myths. A historical relationship un- 
documented by direct evidence is indicated. To assign the myth 
itself to a definite period is far more difficult. I hesitate to place 
it in the proto-American layer, so long as it remains unknown 
from other equally rude peoples such as the Brazilian Ge. There 
is likewise the question of stability. How probable is it that a 
tale would retain essential elements for, say, five thousand years? 
Yet Boas has found that quite trivial stories, which at first blush 
would suggest narratives of personal adventures, turn up in identi- 
cal form among widely and long separated branches of the Es- 
kimo. This objection, therefore, is not a serious one. It could also 
be met by Nordenskiold's view that the ancestors of the Fuegians 
traversed South America with great rapidity so as to remain 
virtually unaffected by the influence of the tropical territories 
passed. However, I cannot readily assume that Californians and 
proto-Yaghans lived in amicable neighborliness a few centuries 
ago. Both the "Canoe" and the "Foot" Indians of Tierra del Fuego 
show far too many adaptations to their present habitat to be 
credited with a really recent tenancy. Late diffusion being barred 
by the lack of the myth in intervening areas, I should put its 
origin somewhere in the era intermediate between the occupa- 
tion of the Fuegian archipelago and the specialization of Cali- 
fornian culture. 

The conclusion of historical relations between the Fuegians and 
North American Indians is strengthened by concomitant resem- 
blances. Thus, Father Schmidt points out the rule by which ini- 
tiates in California and in Tierra del Fuego must not scratch 
themselves with their fingers, for which a special little rod is 
substituted.^^ This feature is shared by so many other American 

"Wilhelm Schmidt, Ursprung der Gottesidee, II (1928), 1023. 


Cultural Anthropology: A Science 

groups that it does not by itself demonstrate specific relations 
between the two areas. Nevertheless, its occurrence indicates that 
the mythological parallel does not stand alone. 

As a matter of fact, the use of a head-scratcher merits histori- 
cal consideration in its own right. ^^ This quaint practice ranges 
from British Columbia to the very tip of South America. It is 
inconceivable that such a custom should have spread through 
recent propagandist fervor. Had such been the case, we should 
expect to find it in the same invariable setting, whereas actually 
the situations vary. Often it adheres to puberty initiation, but now 
to that of girls exclusively, now to the boys', and again to that of 
either sex. The implement may be associated with a tribal cere- 
mony like the sun dance or with the homecoming of a defiled war- 
rior, with funeral rites or with the birth of a child. Generalizing 
for the whole hemisphere, we might say that it potentially crops 
up in situations of ceremonial stress or seclusion, but obviously 
a custom is borrowed not in the abstract but in a concrete form. In 
view of the immense distribution of the scratcher, the variations 
in its correlates, and its occurrence among such unequivocally 
simple tribes as the Californians, the Ge, and the Fuegians, I 
think we can safely assign it to an extremely early layer of Ameri- 
can culture, to the same period roughly as stone-boiling, to an 
earlier period than the Marplot theme. 

Ethnology may thus engage in chronological hypotheses, which 
are not indeed experimentally verifiable, but, like the geologist's 
and zoologist's historical conjectures, may become more probable 
in the light of extraneous facts. The stratigraphic findings of 
archaeology claim the validity of any direct scientific demonstra- 

Science, however, takes cognizance not only of the categories 
of space and time but also of causality. I am acquainted with the 
philosophical problems connected with this matter since the days 

" Edward Sapir, "A Girl's Puberty Ceremony among the Nootka Indians," 
Transactions of the Royal Society of Canada, 3d ser., VII (1913), 79; Cora Du 
Bois, Wintu Ethnography, Univ. Calif. Publ. Amer. Arch, and Ethn., XXXVI 
(1935), 52; A. L. Kroeber, Handbook of the Indians of California, BAE, Bull. 78 
(1925), p. 254; Leslie Spier, Yuman Tribes of the Gila River (Chicago, 1933), 
pp. 180, 325, 327; Th. Koch-Grunberg, Vom Roroima zum Orinoco (Stuttgart, 
1923), III, 130, 168; communication from Curt Nimuendaju about the Timbira. 


Theories and Theorists 

of Hume and Kant, and the attempts to supplant causality with 
the mathematical notion of function by Mach and Bertrand Rus- 
sell are familiar to me. For my present purposes the distinction, 
however, seems unessential. The question is simply whether eth- 
nology can give satisfactory reasons why cultural phenomena are 
as they are. To a certain extent the answer has already been 
given. Every ecological interpretation involves at least one defi- 
nite determinant of the fact to be accounted for. The Chaco tribes 
ceased to make stone tools in their present habitat because they 
no longer had the raw material; Australians do not grind stone 
when the diorite suitable for such a technique is beyond reach, 
etc. Explanations of equal simplicity often emerge when the rami- 
fying branches of a once united stock develop different traits. An 
oft-cited illustration is that of the Athabaskan family, divisible 
into one branch that extends from Alaska to Hudson Bay; another 
in and near north westernmost California; and a third in New 
Mexico and Arizona. The Navaho of Arizona, for example, have 
quite obviously taken over features of Pueblo life such as maize 
cultivation, which clearly set them off from their congeners in 
the north; and the Hupa of California have become almost in- 
distinguishable from their alien neighbors. Were it not for the 
testimony of language, indeed, no one would ever suspect either 
Navaho or Hupa of Athabaskan affinities. Similarly, the Chiri- 
guana, who emigrated from Paraguay toward the Andean region 
in the sixteenth century, have adopted various features distin- 
guishing them from their fellow-Guarani.^^ 

Logically, this type of problem presents itself as follows. Group 
A, which once formed part of a major unit, has become detached 
and now differs from Groups B and C; how are the differences 
to be explained? If the remote segments came to vary in the same 
direction, we might think of a common law of development; since 
they diverge, the differences must be due either to independent 
evolution or to novel influences from new neighbors. In some 
instances, such as the adoption of a hitherto unknown plant not 
available in its wild form, only the second alternative is possible; 
and in many other cases it ojBFers the only probable explanation. 
This is not evading the above-mentioned problem of selective 

^^ Nordenskiold, op. cit., VIII (1930), 157. , ■ 


Cultural Anthropology: A Science 

borrowing. For, in our Southwest the Navaho or Maricopa have 
not taken over the architecture of the Pueblo tribes; while in the 
Chaco the Choroti and the Chiriguano represent a comparably 
lower and higher plane of living. Contact with new groups is no 
master-key for all hidden mysteries; suffice it that it solves defi- 
nitely and definitively a particular set of problems. 

Some minds, however, are not content with so specific a form 
of interpretation, but aspire to principles of absolute generality. 
Reverting to what I have already said, I insist that this ideal is 
only to a most moderate degree realized even in physics while 
the saner philosophers of science have increasingly deprecated 
possession of a Laplacean formula for all the past, present, and 
future happenings in the universe. In the study of culture we can 
certainly put forth universally valid propositions, but they are not 
particularly interesting. It is true that not a single hunting tribe 
has mastered analytical geometry, but somehow the statement, 
impeccable though it be, fails to shed illumination. On the other 
hand, the "law" once postulated that the rituals of hunters-gath- 
erers revolve about food species is not a law at all. Unquestionably 
many ceremonials are devoted to insuring the food supply. But 
the buffalo-hunting Crow had as their main rituals the sun dance 
and the tobacco dance, the former pledged to secure vengeance, 
the latter for the generic benefit of the tribe. In part of California 
the mourning anniversary embodies the peak of native ritualism; 
in some other sections the girls' adolescence ceremony looms large. 
A real law would have to account for such deviations. 

The social scientist who plays the sedulous ape to mechanics 
makes himself ridiculous at a time when the hegemony of me- 
chanics is not even recognized in other departments of physics or 
chemistry. The social scientist, like the organic chemist, is not 
obliged to conform to a model set by an alien discipline; his busi- 
ness is to coordinate in consonance with the nature of his phe- 
nomena. Now to date these phenomena suggest to me no absolute 
laws; but they do point toward certain regularities, and these it is 
certainly our duty to ascertain as rigorously as possible. 

In our central states a solid block of Algonkian (Menomini, 
etc.) and Siouan (Omaha, etc.) tribes share the so-called "Omaha 
type" of nomenclature, i.e., they call the son of a maternal uncle 


Theories and Theorists 

by the same term as the uncle himself. In this respect they differ 
conspicuously from fellow-Algonkian and fellow-Siouan tribes, 
respectively. This divergence can, of course, be explained by bor- 
rowing along the lines indicated; if, say, the Algonkians were the 
first to adopt this mode of classing certain cousins, then their 
Siouan neighbors differ from other Siouans because of the Algon- 
kian influence. But obviously the explanation cannot simulta- 
neously serve for both deviating groups; so on this assumption it 
remains unexplained why the Algonkians of the area should have 
departed from prior Algonkian usage. To discover the determi- 
nants of the phenomenon discussed, we naturally inquire whether 
it occurs elsewhere; and, if so, with what concomitants. The ge- 
ographically nearest parallel to the Omaha type exists among the 
Californian Miwok and their neighbors. Any historical relation 
between the Miwok and the Indians of our central states is out of 
the question; fifteen hundred miles apart, they are separated by 
tribes with no trace of an intermediate kinship system, without 
a suggestion of a linguistic affiliation with either Algonkian or 
Siouan. What, then, causes the strange duplication of failing to 
distinguish between two adjacent generations, between an uncle 
and a cousin? An obvious feature common to the central Al- 
gonkians, southern Siouans, and Miwok is their patrilineal clan 
system, for such an organization puts the uncle and nephew in 
question in the same clan, their classification under one head be- 
coming intelligible: the clan bond overrides such factors as age 
and generation. But this is not an adequate explanation, be- 
cause other patrilineal tribes, such as the Algonkian Ojibwa, lack 
the Omaha nomenclature, i.e., distinguish the avuncular-nepotic 
generations. If patrilineal descent is involved at all, then, it must 
be reinforced by a supplementary factor. Now, the Omaha and 
the Miwok share a marriage rule that tends to level the difference 
of generations in question. Both tribes permit a man to marry his 
wife's niece, i.e., her brother's daughter; the wife's niece is thereby 
equated with the wife, and so are their respective children, inas- 
much as they are all equally the children of the same man. Let 
us, then, take the son of the niece in relation to that niece's brother; 
the older man is literally the maternal uncle of the younger. But 
the double marriage makes the first wife also a "mother" of her 


Cultural Anthropology: A Science 

niece's children since she, too, is a wife of their father. Hence, 
the brother of the older wife is another "maternal uncle" in prin- 
ciple, while actually he is of course the father of the true uncle. 
Hence, the same person is simultaneously maternal uncle and ma- 
ternal uncle's son to the offspring of the younger wife (his sis- 
ter). The mysterious confounding of generations is explained by 
the sanctioning of simultaneous or successive marriage with a 
woman and her brother's daughter. 



The provisional conclusion, then, is that the Omaha peculiarity 
is a function of patrilineal descent and a certain rule of marriage. 
It is quite possible that additional determinants must be brought 
in to explain all occurrences of the phenomenon. It is even pos- 
sible that other determinants may yield the same result. What I 
am concerned with is the method of procedure, the recognition 
of how an explanation may be reached. That the same effect may 
be due to convergence is true and must be accepted as a fact 
wherever it is demonstrable. But this never explains the resem- 
blance to be accounted for, since we possess no organ for grasping 
how unlike causes can yield like effects, only a principle of 


Theories and Theorists 

causality by which the same causes are understood to produce Kke 
effects. We have the choice of accepting the incomprehensible 
with a shrug over the complexity of our data; or of seeking the 
uniform antecedents of similar facts. In the latter case we can 
proceed only by the approved processes of logic, the method of 
varying concomitants. Only by this procedure shall we be able 
to corroborate or repudiate the numerous correlations that are be- 
ing constantly alleged and so rarely demonstrated. I am confident 
that from a thoroughgoing application of this method a number 
of significant regularities will emerge, representative of the type 
of generalization consistent with the nature of our field at the con- 
temporary stage of insight. Our generalizations will not assume 
the form of a Laplacean formula but the form of propositions such 
as these: The Omaha nomenclature is a function of patrilineal 
descent when coupled with certain intergeneration marriages; 
the avunculate is a function of matrilocal residence coupled with 
such-and-such further conditions; handmade pottery is a woman's 
craft except when it becomes a specialized economic activity. 

Ethnology will make immeasurable strides forward as soon as 
everyone who affirms correlations will favor his colleagues with 
a proof on an inductive basis. 

To sum up: Ethnology is simply science grappling with the 
phenomena segregated from the remainder of the universe as 
"cultural." It is a wholly objective discipline, whether it deals with 
subjective attitudes or not, for its function is the determination of 
reality in verifiable terms. It coordinates its data spatially, in so 
far forth duplicating the procedure of geography. It coordinates 
its data chronologically to that extent sharing the logic of geology, 
paleontology, historical astronomy, and political history; the par- 
ticular techniques employed must vary with the problem, as in 
other branches of learning. Finally, it coordinates in terms of 
causality as the concept has been epistemologically purified; and 
by the demonstration of functional relationships it may attain 
the degree of generalization consistent with its own section of the 



Evolution in Cultural Anthropology: 
A Reply to Leslie White 

Leslie White's last three articles in the American 
Anthropologist ^ require a reply since in my opinion they obscure 
vital issues. Grave matters, he clamors, are at stake. Obscurantists 
are plotting to defame Lewis H. Morgan and to undermine the 
theory of evolution. 

Professor White should relax. There are no underground 
machinations. Evolution as a scientific doctrine — not as a farrago 
of immature metaphysical notions — is secure. Morgan's place in 
the history of anthropology will turn out to be what he deserves, 
for, as Dr. Johnson said, no man is ever written down except by 
himself. These articles by White raise important questions. As 
a victim of his polemical shafts I should like to clarify the issues 
involved. I premise that I am peculiarly fitted to enter sym- 
pathetically into my critic's frame of mind, for at one time I was 

American Anthropologist, XLVIII (April-June, 1946), 223-233. 

^ Leslie A. White, "Energy and the Evolution of Culture," American Anthro- 
pologist, XLV (1943), 335-356; idem, "Morgan's Attitude toward Religion and 
Science," ibid., XLVI (1944), 218-230; idem, "'Diffusion vs. Evolution': An 
Anti-Evolutionist Fallacy," ibid., XLVII (1945), 339-356. 


Theories and Theorists 

as devoted to Ernst Haeckel as White is to Morgan. Haeckel had 
solved the riddles of the universe for me. 

Considering the fate of many scientific men at the hands of their 
critics, it does not appear that Morgan has fared so badly. Ameri- 
cans bestowed on him the highest honors during his lifetime, 
eminent European scholars held him in esteem. Subsequently, 
as happens with most celebrities — Aristotle, Darwin, George 
Ehot, for example — the pendulum swung in the opposite direc- 
tion. The reaction overshot its mark at times, as when American- 
ists doubted even Morgan's Crow findings. Nevertheless, appreci- 
ation has been frequent and ample even in later periods. Haddon 
calls Morgan "the greatest sociologist of the past century"; 
Rivers hails him as the discoverer of the classificatory system; 
Radcliffe-Brown rates Systems of Consanguinity and Affinity a 
monument of scholarly, patient research; Mitra pictures Morgan 
as an anthropological colossus, a greater Tylor; Marcel Mauss and 
Paul Radin are avowed admirers.^ What, precisely, does White 
expect? An academic muezzin at every center of learning who 
shall lead anthropologists in daily Rochester-ward obeisances 
and genuflections? 

The term "Boasian" is misleading. Of the great physiologist 
Johannes Miiller a one-time disciple said: "There is no school in 
the sense of common dogmas, for he taught none, only a common 
method." This holds for Boas. His students have often differed 
from their teacher and from one another. Kroeber, Sapir, Radin 
have repeatedly expressed their dissent from cardinal "Boasian" 
views, and even I have uttered misgivings on certain points. 
Laufer incidentally, was not trained by Boas at all and was doubt- 
less more deeply influenced by Eduard Hahn. 

=" Alfred C. Haddon, History of Anthropology (London, 1934), p. 127; W. H. R. 
Rivers, Kinship and Social Organisation (London, 1914), pp. 4f.; A. R. Radcliffe- 
Brown, "The Study of Kinship Systems," RAI, Journal, LXXI (1941), 4 f.; 
Panchanan Mitra, A History of American Anthropology (Calcutta, 1933), pp. 


Evolution in Cultural Anthropology 

Speaking pro dome, I find White's procedure curious. He vir- 
tually accuses me of plagiarizing Morgan on the subject of animal 
domestication ^ when I merely state matters of long-established 
common knowledge. Morgan's reference, incidentally, is so casual 
as hardly to merit notice, and Francis Galton's full discussion of 
the point, which I have duly registered, is much earlier. Again, in 
White's latest article I am referred to fifteen times, but only three 
of the publications cited appeared after 1922, though I have twice 
dealt with Morgan rather fully in much later years.* In these re- 
cent discussions, as in my Primitive Society long ago, I explicitly 
mention Morgan's use of diffusion,^ yet White finds it "difficult to 
see how Lowie could have read the passages in Morgan" con- 
cerned with that principle. 

I do not, however, impugn White's good faith; the obsessive 
power of fanaticism unconsciously warps one's vision. 

As a matter of fact, my conscience is clear on the subject of 
Morgan. In 1912, in the face of Americanist skepticism, I sub- 
stantiated his discovery of matrilineal clans among the Crow. In 
1916 I referred to his "superb pioneer achievements"; in 1917 I 
took pains to show that he was right and I wrong on an important 
point in Crow kinship nomenclature. In 1920 I called Ancient 
Society "an important pioneer efi^ort by a man of estimable in- 
telligence and exemplary industry"; in 1936 I commended his 
acuity as a field worker and credited his Systems with "a magnifi- 
cent and valid conception." ^ 

To be sure, eulogistic comments are balanced by harshly criti- 

* White, "Energy and the Evolution of Culture," p. 339. 

* R. H. Lowie, "Lewis H. Morgan in Historical Perspective," in Essays in Anthro- 
pology Presented to A. L. Kroeber (Berkeley, 1936), reprinted as No. 26 in 
the present volume; idem, The History of Ethnological Theory (New York, 1937). 

^ Idem, The History of Ethnological Theory, p. 59; idem. Primitive Society 
(New York, 1937), p. 147. 

" Idem, Social Life of the Crow Indians, AMNH, Anthropological Papers, IX 
(1912), p. 186; idem, "Historical and Sociological Interpretations of Kinship 
Terminologies," in Holmes Anniversary Volume (Washington, 1916), p. 293 
(reprinted as No. 3 in the present volume); idem. Notes on the Social Organiza- 
tion and Customs of the Mandan, Hidatsa, and Crow Indians, AMNH, Anthropo- 
logical Papers, XXI (1917), 56; idem. Primitive Society (New York, 1920), p. v; 
idem, "Lewis H. Morgan in Historical Perspective," pp. 170, 180; idem. The His- 
tory of Ethnological Theory, p. 92. 


Theories and Theorists 

cal ones. As my treatment of Boas indicates/ I am not an idolater. 
In proof of my malevolence White likes to quote a sentence dating 
back to 1920: "It may be said categorically that even at his worst 
Morgan never perpetrated more palpable nonsense, and that is 
saying a good deal." ^ More suo. White fails to explain what evoked 
such violence. Since in 1877 there were already trustworthy re- 
ports on African Negroes and Polynesians, I regarded Morgan's 
denial of monarchic and aristocratic institutions among primitive 
peoples as inexcusable. Nowadays I should use more temperate 
phraseology, but as to the substance of my remark I remain ada- 
mant. Indeed, by way of amendment, I should say that Morgan 
is guilty of still less defensible propositions. How could any eth- 
nographer ever put the Polynesians into the same category with 
Australians and below the Northern Athabaskans? Again, Morgan 
finds that "the discrepancies between them [Seneca and Dravidian 
kinship systems] are actually less . . . than between the Seneca 
and the Cayuga." This does not deter him from inferring a racial 
affinity between Dravidians and Seneca because of the identity 
of their kinship terminologies.^ The Seneca, by implication then, 
must be racially closer to the Tamil than to their fellow-Iroquois. 
Is this line of argument to be rated brilliant, profound, sensible, 
dubious, or is it palpable nonsense? How does Professor White 
grade it? 

Professor White may say that we ought to judge a scholar by 
his positive contributions, and I heartily concur. Yet fairness to 
other scholars viewed in historical perspective demands that we 
should not gloss over such flagrant delinquencies, especially when 
their perpetrator regards them as cardinal discoveries. I should 
like to see some realization on White's part that sporadic im- 
patience with Morgan may have an objective basis. 

Certainly irritation at him is not necessarily bound up with 
anti-evolutionism. How, otherwise, does White explain the gen- 
erally tender treatment of Tylor? I herewith offer some purely 

''Idem, The History of Ethnological Theory, pp. 151-155; idem, "Franz Boas, 
His Predecessors and His Contemporaries," Science, XCVII (1943), 202-203. 

® Idem, Primitive Society, p. 389. 

® L. H. Morgan, Systems of Consanguinity and Affinity of the Human Family 
(V\^ashington, 1871), pp. 166, 508. 


Evolution in Cultural Anthropology 

personal remarks to explain why, notwithstanding my appreciation 
of certain aspects of Morgan's work, I cannot take kindly to him 
as a scientific personality. For one thing, I resent his dreary 
schematism; for another, I find little evidence in him for that 
sympathetic projection into alien mentality which anthropology 
is supposed to foster. He avoids one of the major departments 
of culture because "all primitive religions are grotesque and to 
some extent unintelligible." ^*^ He dogmatizes thus on the subjec- 
tive feelings of our indigenes: "The passion of love was unknown 
amongst the North American aborigines of pure blood. The fact 
is sujfficiently established by their marriage customs. They were 
given in marriage without being consulted, and often to entire 
strangers." ^^ Were, however, any doubt possible as to Morgan's 
narrow-mindedness, it is dispelled by White's admirable edition 
of his hero's travel notes. A few gems suffice for demonstration: 
The frescoes of Michael Angelo are "substantially absurd," the 
Sistine Chapel is "a poor specimen of a Pagan Temple." The 
Roman Carnival is "an unutterable piece of nonsense and levity," 
proving the frivolity of the population. Catholic ceremonial evokes 
"inexpressible disgust." The people of southern Italy "are utterly 
worthless," Italians in general "degraded beyond all other peoples 
called civilized." On the other hand we learn: "Our country is 
the favored and the blessed land. Our institutions are unrivalled 
and our people the most advanced in intelligence. . . ." ^^ 

Tylor, I feel, could not have written in this strain. Probably 
White does not acquiesce in his idol's Bilboesque sentiments. If 
this is a correct surmise, why does he dissent from Radcliffe- 
Brown's apt characterization of Morgan's provincialism? ^^ Why 
does he ascribe criticism of Morgan to anti-evolutionism or other 
sinister motives when such obvious reasons are at hand? Evolu- 
tion has very little to do with the case. 

^° Idem, Ancient Society (Kerr ed.; Chicago, 1877), p. 5. 

" Idem, Systems of Consanguinity, p. 491. 

^ Leslie A. White, ed., Extracts of Lewis Henry Morgan's European Travel 
Journal, Rochester Historical Society Pubhcations, XVI (1937), 285, 290, 303, 
311, 315, 327. 

" Idem, "Morgan's Attitude toward Religion and Science," American Anthro- 
pologist, XLVI (1944), 218 f., 230. 


Theories and Theorists 

As there is no Boasian sect, so there is no Boasian "reactionary 
philosophy of anti-evolution," nor a "philosophy of planless hodge- 
podge-ism." ^^ The former plirase naturally suggests the degenera- 
tion theories which Tylor refuted in Primitive Culture. Contrary 
to what might be regarded as the implications of White's phrases. 
Boas and his disciples nowhere question the established facts 
of prehistory (nor does anyone else), witness his inclusion of 
Nelson's section on "teclinological evolution" in General Anthro- 

It is, indeed, not easy to discover the meaning of Wliite's ac- 
cusations. On the one hand, he tries to clear Morgan of unilinear 
evolutionism, which is the butt of Boas's strictures. If he were 
correct (see below), it would merely prove that Boas misunder- 
stood Morgan, not that he had an anti-evolutionary philosophy. 
On the other hand. White summarizes evolutionary doctrine in his 
"Energy and the Evolution of Culture," laying down propositions 
which Boasians may find trite and futile, but which do not arouse 
the Bryanesque ardor imputed to them when evolution is pro- 
pounded. Boasians do not deny that man requires food, controls 
his environment with the aid of tools, improves his control by 
invention and discovery, and alters social structure as a result of 
technological evolution. I refer White to a paragraph of Boas's 
in his general text.-^^ 

As a matter of fact, no reputable scholar challenges either the 
demonstrable findings of prehistory or the economic truisms pro- 
claimed by White, least of all, the Austrian school whose writ- 
ings are evidently on his Index librorum prohibitorum since he 
contents himself with a garbled sentence borrowed from Kluck- 
hohn concerning their views. It is, however, patent that Fathers 
Schmidt and Koppers are not anti-evolutionists in White's esoteric 
sense. They do not, to be sure, like the word "Evolution," but the 
reality they fully recognize. As Morgan generally speaks about 

^* Idem, "Energy and the Evolution of Culture," p. 355; idem, "Diffusion vs. 
Evolution," p. 354. 

^^ Franz Boas, ed., General Anthropology (New York, 1938), pp. 150 ff. 

'"' Wliite, "Energy and the Evolution of Culture," p. 354; Boas, op. cit., pp. 
678 f . 


Evolution in Cultural Anthropology 

"development," so they have plenty to say concerning Entwick- 
lung, Fortentwicklungen, Weiterbildungen; and when Father 
Schmidt somewhere traces Stufen der ganzen Entwicklung, what 
are these but evolutionary stages? Such section headings as "Der 
Schritt vom niederen zum hoheren Jdgertum" or "Von der Jagd 
zur Tierzucht" indicate that an evolution of some sort is definitely 
assumed within the several Kulturkreise. It is unilinear evolution 
that the Austrians and the Boasians reject, but since White has 
latterly discovered that Tylor and Morgan are not unilinear evolu- 
tionists at all ^^ — a matter to be discussed below — what is the row 
about from White's point of view? It would seem that then "he 
is right and we are right, and all's as well as well can be." If he 
deigned to read the Catholic scholars he so airily dismisses, he 
would discover that in the field of empirical inquiry ( as opposed 
to metaphysics) they are as technologically oriented as himself. 
Stressing economic conditions as von ganz hervorragender Be- 
deutung, they explicitly accept historical materialism as an ex-, 
cellent and even indispensable heuristic principle. ^^ 

Thus White's gloomy picture of most contemporary anthropol- 
ogists plunged into Cimmerian darkness, unrelieved by a single 
lambent ray of evolution, is preposterous. He ought to realize that 
Thurnwald, Radclifi^e-Brown, Radin, Lesser, Malinowski are 
professed evolutionists, and that even I have spoken kindly enough 
of neo-evolutionism.-^^ 

The questions which worry White, viz., "why Boas and his 
disciples have been anti-evolutionists" and what may be "the 
source and basis of the anti-evolutionist philosophy of the Boas 
group," automatically disappear. In order to infuse sense into 
such queries they must be re-formulated : Why have Boas and his 
students attacked not evolution, but Morgan's and other writers' 
evolutionary schemes? 

Characteristically White does not attempt to answer the ques- 

" White, "Diffusion vs. Evolution," p. 347. 

^^ Wilhelm Schmidt and Wilhelm Koppers, Volker und Kulturen ( Regensburg, 
1924), pp. 382, 396 ff., 625 ff., 636. 

" Lowie, The History of Ethnological Theory, pp. 246, 289; Bronislaw Mali- 
nowski, A Scientific Theory of Culture and Other Essays (Chapel Hill, 1944), 
p. 17; Richard Thurnwald, Die menschliche Gesellschaft (Berlin and Leipzig, 
1931-1935), I, 16. 


Theories and Theorists 

tion genetically. He makes a great to-do about Morgan's never 
alleging tlie priority of animal husbandry to agriculture (as if 
anyone had made the charge); he triumphantly points to Mor- 
gan's placing pictography before the alphabet (a matter not in 
dispute ) . But he preserves a discreet silence on virtually all mat- 
ters that are relevant to the debate. 

Boas began as a unilinear evolutionist. In 1888 he defended 
"the current view of a necessary precedence of matrilineal forms 
of family organization." ^^ I also recall his telling me how deeply 
he was impressed by a first reading of Tylor's "adhesion" study. 
All problems then appeared solved, at least in principal. What was 
it, then, that made him alter his convictions? That is a worth- 
while psychological problem. 

Before directly answering this question it is well to digress 
and take up certain startling discoveries announced in White's 
second article, for they explain in part why he does not under- 
stand what the discussion is about. 

Between 1859 and 1881, we learn, thinking people were divided 
into two hostile camps: they championed either science or the- 
ology. A devout Christian could not be a scientist or a Darwinian. 
"Those who opposed DarwinisTn did not labor for, or make con- 
tributions to, science. ... If you were for Theology, you were 
against Science." ^^ 

This statement happens to be wrong in eveiy particular and 
from every conceivable angle. Darwinism and science never have 
been interchangeable terms. Pious Catholics, witness Pasteur and 
Mendel, made epoch-making researches during Morgan's life- 
time; Clerk Maxwell and Kelvin were, I believe, devout Protes- 
tants; Julius Robert Mayer was beyond any doubt a deeply 
religious man — and an opponent of Darwinism.-- The critics of 
Darwinism included towering figures in the history of science — 
Karl Ernst von Baer, Louis Agassiz, Rudolf Virchow, Albert Kol- 

^ Franz Boas, Race, Language and Culture ( New York, 1940 ) , p. 635. 
^ White, "Morgan's Attitude," p. 219; the itahcs are Wliite's. 
^Wilhelm Ostwald, Grosse Manner (Leipzig, 1910), pp. 73, 82. 


Evolution in Cultural Anthropology 

liker, Sir Richard Owen. Some of them were religious, others mere 
skeptics. On the other hand, Christian behevers by no means 
uniformly rejected evolution; they included the geologist Joseph 
Le Conte and the botanist Asa Gray, whom Darwin held in high 

The Manichaean picture of Darwinian atheists as angels of 
light pitted against a Satanic brood of Christian obscurantists is 
merely more of Whitean melodrama. As might be expected, so 
revolutionary a doctrine as Darwinism evoked a variety of re- 
sponses. Darwin's, Wallace's, Huxley's correspondence and Haeck- 
el's polemical writings furnish ample illustrations. True scientists 
were thrilled over having so many obscurities illuminated for 
the first time. This group, as noted, comprised Christians, who 
promptly set about to harmonize their religion with their scien- 
tific convictions, revising what seemed unessential articles of faith. 
Materialists and other radicals just as naturally used the new ideas 
as grist for their mill, yet not all of them did so undiscriminat- 
ingly. David Friedrich Strauss, e.g., hailed Darwinism as a great 
achievement, yet found it still "highly imperfect." The inade- 
quacies felt even by so sympathetic a scholar were naturally ag- 
grandized and stressed by fundamentalists. Thus, there was by no 
means a clear-cut division into two hostile camps. 

For present purposes we are interested in the scientific opposi- 
tion. Why, we ask, did a man like Virchow maintain so reserved 
and at times hostile an attitude towards Darwinism? Why, half a 
century later, did experimentalists like Thomas Hunt Morgan re- 
main critical? Whence the skepticism of Jacques Loeb,^^ whom 
even White can hardly brand as a fundamentalist? Were all these 
men reactionary philosophers, enemies of Science? The answer 
is clear. They objected to evolutionary theories on the ground that 
they were not scientific enough. Some of them cannot, neverthe- 
less, qualify as "anti-evolutionists," for they gratefully recognized 
the widening of the intellectual horizon due to Darwin. But they 
rejected phylogenetic speculation, for which they sought to sub- 
stitute the demonstrable findings of the laboratory. 

^ Jacques Loeb, in Heinrich Schmidt, ed., Was wir Ernst Haeckel verdonkcn 
(Leipzig, 1914), II, 15. 


Theories and Theorists 

Reverting to Boas, his critique of evolutionary schemes is the psy- 
chological equivalent of the experimentalists' critique of "the bio- 
genetic law." The facts did not fit the theory, hence the theory 
would have to be modified or discarded. To cite concrete instances, 
L. H. Morgan teaches that the individual family is an end-product, 
preceded by various stages including that of a clan organization; 
Morgan, Bachofen, and Tylor teach the priority of matrilineal 
descent. Boas found that in the interior of British Columbia clan- 
less tribes with a family organization and a patrilineal trend 
adopted from coastal neighbors a matrilineal clan organization."^ 
Diff^usion thus disproved the universal validity of the formula that 
Boas himself had been defending in 1888. 

Subsequently it turned out that borrowing had played a far 
greater part among primitive groups than most anthropologists 
had supposed. The difiiculty of establishing ujiiversal laws of 
sequence seemed correspondingly increased. In this sense T came 
to maintain that difi^usion laid the axe to the root of any theory of 
historical laws. 

Professor White in dissent blares forth a sennet of defiance: ( 1 ) 
Tylor and Morgan both accept diffusion in concrete cases, hence 
Diffusion and Evolution lie together as the leopard and the kid in 
Isaiah's peaceful kingdom. (2) Tylor "does not state, nor do his 
remarks imply or even allow of the intimation 'that every people 
must pass through all the stages of development,' as Boas claims." 
Similarly does Morgan nowhere "declare or even imply that 
each tribe, everywhere, must go through the same stages of cul- 
tural development."-^ (3) The Boasians fail to discriminate be- 
tween cultural evolution and the culture history of specific tribes 
or peoples; Tylor and Morgan are never concerned with the his- 
tory of tribes or peoples, only with that of cultural traits or com- 
plexes, such as writing, metallurgy, social organization. Hence the 

^* Franz Boas, Die Resultate der Jesup-Expedition ( Separat-Abdnick aus den 
Verhandlungen des XVI Internationalen Americanisten-Kongresses, Wien, 1909), 
p. 16; John R. Swanton, "A Reconstruction of tlie Theory of Social Organiza- 
tion," Boas Anniversary Volume (New York, 1906), p. 173. 

^ White, "Diffusion vs. Evolution," p. 347. 



Evolution in Cultural Anthropology 

criticism that their formulae do not fit particular tribes is irrele- 
vant, they were never meant to do so. 

Let us scrutinize these allegations. That Tylor and Morgan 
knew about diflFusion is, indeed, a patent fact, which I have 
taught for over twenty-five years, as White admits with reference 
to Tylor.^^ The point is irrelevant, for it is failure to integrate dif- 
fusion with evolution that is charged. To repeat a twice-told 
tale, Tylor offers the formula: (1) patrilocal residence; (2) taboo 
between wife and husband's kin. If the correlation in a dozen cases 
evolved independently, there is presumably an organic tie-up. But 
what if there has been a single historic center of origin for the 
trait couple? As a transcendent entity there might still be a law 
of sequence, but there would be no way of demonstrating it. Simi- 
larly with Morgan. If clans arise independently the world over 
out of similar antecedents, the clan may reasonably be put into a 
definite place in a chronological series. Not so if, as Morgan argues, 
it sprang up a single time and was thus diffused over the globe. 
Metaphysically, here also the sequence postulated is conceivable; 
empirically, however, it ceases to be demonstrable. To amend 
White's phraseology, "diffusion negates [the possibility of prov- 
ing] evolution." 

The criticism is that Tylor and Morgan fail to resolve this 
logical difficulty, not that they ascribed an independent develop- 
ment of Christianity to the Seneca or believed in an independent 
origin of maize-growing in the Balkans. 

Contrary to White's allegations, moreover, Tylor and Morgan 
much more than imply a faith in parallelism. Both strongly be- 
lieved in psychic unity. In accepting this principle Morgan de- 
clares : "It was in virtue of this that mankind were able to produce 
in similar conditions the same implements and utensils, the same 
inventions, and to develop similar institutions from the same origi- 
nal germs of thought." And in one of his most famous publica- 
tions Tylor, having likened human institutions to stratified rocks, 
thus continues: "They succeed each other in series substantially 
uniform over the globe independent of what seem the compara- 

"" Lowie, Primitive Society, p. 147; idem, "Lewis H. Morgan in Historical Per- 
spective," p. 173. 



Theories and Theorists 

lively superficial differences of race and language, but shaped 
by similar human nature. . . ." ^^ Now, this last statement, to 
be sure, says nothing about specific peoples. The important thing, 
however, is not how a writer formulates his principles in the 
abstract but what he does with them. 

Now, White to the contrary notwithstanding, Tylor and Morgan 
both apply their magical formulae to particular peoples. Tylor 
correlates cross-cousin marriage with exogamy. His schedules 
listed 21 peoples as practising cross-cousin marriage, only 15 of 
/ whom were described as exogamous. Sure of the general validity 
of the formula, he boldly places the remaining six tribes in the 
exogamous category. 

Morgan finds Ponca, Winnebago, Ojibwa, and Menomini with 
patrilineal institutions, but from his formula he infers that they 
were once matrilineal. Similarly with the Greeks, despite "the 
absence of direct proof of ancient descent in the female line." 
Correspondingly, Greeks, Romans, Celts, Germans and Hebrews 
are credited with a pristine punaluan stage.^^ 

I suggest that this is no longer dealing with the cultural process, 
it is veritably "psuedo-history." Morgan is speaking of the past 
of Siouan and Algonkian tribes, and he defines it not on anything 
known about Siouan and Algonkian history, but as a deduction 
from his formula. And when a punaluan stage is ascribed to the 
remote ancestors of Greeks and Hebrews, it can be only on the 
assumption that these peoples have passed through the same stages 
as other peoples. Morgan does not ask in how far unique hap- 
penings in the past — say, alien influences — might have deflected 
these peoples from their predestined path. 

Incidentally, White nowhere explains how he supposes the 
formulae to have been ultimately derived. Are they empirical 
inductions? In that case they must rest on observations of the his- 
tory of specific tribes. Or are they all a priori constructs like the 
precious notion about the uncertainty of fatherhood as the cause 
of matrilineal reckoning in early times? 

-"^Morgan, Ancient Society (Kerr ed. ), p. 562; E. B. Tylor, "On a Method of 
Investigation of the Development of Institutions; Applied to Laws of Marriage 
and Descent," RAI, Journal, XVIII (1889), 245-269. 

^ Tylor, op. cit.; Morgan, Ancient Society, pp. 48 f., 59, 159, 161, 170, 175, 357, 
388 f., 438. 


Evolution in Cultural Anthropology 

Leslie White misunderstands the status of the problem. It is false 
that any reputable anthropologist nowadays professes an anti- 
evolutionist philosophy in the sense alleged. The "anti-evolution- 
ism" of the Boasians and of the Kulturkreislehrer has nothing to 
do with, say, the degeneration theory of de Maistre. It implies bad 
faith or bigotry to suppose that either Boas or Schmidt denies 
the findings of prehistory concerning the sequence of stone and 
metal tools, of simpler and of more complex economic systems. 

But the Boasians do claim the right to check evolutionary gen- 
eralizations by the facts they are meant to explain, precisely as 
the experimental zoologists checked the recapitulation theory. A 
dogmatist naturally cannot understand that true scientists are 
not interested in proving a preconceived system. There are such. 
Psychologically viewed, Boas's attitude is simply that of Vir- 
chow towards Haeckel's phylogenetic hypotheses, that of Loeb in 
preferring demonstrable truths to fictitious genealogical trees. 
It is of a piece with Newton's "Hypotheses nan jingo"; with Mach's 
aim to purge science of metaphysics; with Virchow's, Ostwald's, 
Mach's prescription to observe without preconceptions. 

In Virchow's valedictory speech at Wiirzburg in 1856, so young 
Haeckel reported to his parents, the great pathologist explained 
to the students that his whole life's aim was devoted to discover- 
ing the unvarnished truth, to recognize it free from bias and to 
disseminate it unaltered. He exhorted them to get rid of all prej- 
udices, "with which we are unfortunately crammed full from 
infancy on "(mit denen wir leider von Kind auf an so vollgepfropft 
werden) and to view things as they really are ("die Dinge so ein- 
fach und natiirlich anzusehen, wie sie sind")?^ Some years pre- 
viously, in one of his most famous papers, Virchow had already 
warned against the dangers that lurk in any system: 

"Dann kommt jeden Augenblick der Conflikt zwischen dem 
System und dem einzelnen Fall, und gewohnlich wird der einzelne 
Fall dem System geopfert." ^^ 

^^ Ernst Haeckel, Entwicklungsgeschichte einer Jugend (Leipzig, 1921), pp. 
200 f. 

"" Rudolph Virchow, Die Einheitsbestrebungen in der wissenschaftlichen Medicin 
(Berlin, 1849), p. 18. 


/ Theories and Theorists 

Boas, too, refuses to sacrifice individual observations to a pre- 
conceived scheme, voild tout. 

In conclusion I reiterate that I am altogether convinced of 
White's good faith. But as Voltaire explains: 

"La chose la plus rare est de joindre a raison avec L'enthou- 
siasme: la raison consiste a voir toujours les choses comme elles 
sont. Celui qui dans I'ivresse voit les objets doubles est alors prive 
de la raison." 

I have just encountered inexplicably forgotten passages in Father 
Schmidt's Handbuch der Methode der kulturhistorischen Eth- 
nologic (Miinster, 1937), which demonstrate beyond a doubt 
that he accepts not only the concept, but even the term "evolu- 

Nun, Professor Lowie und andere werden aus meiner Darstellung 
der Methode ersehen, dass ich nicht nur das Wort, sondern auch 
den Begriff und die Tatsache der Evolution nicht vermeide, sondern, 
mit der ganzen kulturhistorischen Schule, frank und frei mich zur 
Evolution bekenne, aber nach wie vor in der Ablehnung des Evolu- 
tionismus verharre (p. vii). 

. . . wer den Evolutionismus bekampft und ablehnt, bekampft 
und verwirft damit nicht die Evolution, die (innere) Entwicklung 
(p. 10). 

The case against White is thus even stronger than previously in- 



Franz Boas, 1858-1942 

Franz Boas, for many years the undisputed dean of 
American anthropologists,^ was born in Minden, Westphalia, on 
July 9, 1858. The son of educated parents in easy circumstances, 
he enjoyed standard preparatory instruction; and to the high ethi- 
cal teaching imbibed in the household he referred feelingly in an 
open letter to President Von Hindenburg ( March 27, 1933 ) . 

Entering the University of Heidelberg in 1877, he later shifted 
to Bonn and ultimately to Kiel, where he took his Ph.D. in 1881. 
Though his major interests then lay in physics and geography — his 
dissertation dealt with the recognition of the color of water — , his 
principal professor, Theobald Fischer, directed him also towards 
the historical and ethnographic aspects of geography. Through the 
mathematical training acquired during his university days Boas 
was subsequently able to follow, critically and constructively, 

National Academy of Sciences, Biographical Memoirs, XXIV, No. 9 (1947), 

^ The American Anthropological Association has issued an obituary memoir 
(Memoir 61, 1943), which contains a complete bibliography and articles on Boas 
as a man by A. L. Kroeber, as an ethnologist by Rutli Benedict, as a linguist by 
M. B. Emeneau, as a physical anthropologist by Melville J. Herskovits, as a folk- 
lorist by Gladys A. Reichard, and as an archeologist by J. Alden Mason. I have 
drawTi upon tliis publication, especially on Kroeber's biographical essay. 


Theories and Theorists 

the rise of biometrics and its anthropological applications. But 
he did not narrowly specialize. For example, he read Gustav 
Theodor Fechner, including the delightfully humorous Vier 
Paradoxa. From his son-in-law, Dr. Cecil Yampolsky, I learn 
that Boas's letters of this period have been carefully preserved and 
that they reveal the nascent investigator's ardor for research. 
Publication of the correspondence would be a great boon, for it 
is hkely to reveal intimate glimpses of the writer's personality, 
such as are all too rarely vouchsafed by his monographs and books. 

Two years after the doctorate came the crucial expedition to 
BafRnland, ostensibly in the interests of geographical exploration, 
but ushering in a new era in Boas's life and in the history of 
Eskimo ethnography. Homeward bound. Boas paid his first visit 
to the United States and to New York. On his return to Germany 
he attached himself as an assistant to the Konigliches Museum 
fiir Volkerkunde in Berlin, the institution founded and headed by 
Adolph Bastian; and in 1886 he received peiTnission to lecture at 
the University as a docent. Doubtless he regularly attended the 
Berlin Society for Anthropology, Ethnology, and Prehistory, meet- 
ing its ruling spirit, the great pathologist Rudolf Virchow, to 
whom years later Boas paid a glowing obituary tribute.^ It would 
be most interesting to know how these congenial intellects in- 
teracted, but nothing specific on their relationship is known. The 
obvious similarities between the two men have been repeatedly 
noted — their keen, analytical powers, their exceptional capacities 
for varied work, their independence, moral courage, and alert 
social consciousness. Similarly, it would be worth knowing to 
what extent Bastian influenced the younger man. Conceivably 
Boas's insistence on definite proof of cultural diffusion goes back 
to this source, but it is quite as plausibly explained in terms of 
Boas's own mentality. What personal intercourse with the older 
man doubtless did provide was an intimate knowledge of Bastian's 
theoretical views, often veiled for the mere reader by the most 
crabbed of styles. 

The expedition to Baffinland yielded a number of papers, both 
popular and technical, on the region, tlie ethnographic publica- 

' Science, n.s., XVI (1902), 441-445. 


Franz Boas, 1858-1942 

tions culminating in the monograph on "The Central Eskimo." ^ 
In the meantime Boas found a new realm to conquer. A party 
of Bella Coola Indians exhibited in Berlin and the ample collec- 
tions of the Museum stimulated an interest in northwestern North 
America. Boas pumped the natives for linguistic information, 
published the data secured, and in 1886 himself set forth for the 
coast of British Columbia. Thus started a notable research pro- 
gramme that occupied him literally until his death. 

Returning to New York in 1887, Boas accepted a position as 
Assistant Editor of Science and married Marie Krackowizer, the 
daughter of an Austrian physician and forty-eighter who had 
gained distinction in America both as a medical man and a polit- 
ical reformer. Henceforth the United States became Boas's home. 

The British Association for the Advancement of Science had 
created a committee for the study of the tribes of British Colum- 
bia. From 1888 on, during Edward B. Tylor's chairmanship, Boas 
repeatedly revisited the Northwest coast under this body's aus- 
pices. His early reports bear witness to the range of his interests, 
which took in not only ethnography, but also linguistics and 
somatology. Sometime during these years Boas visited Tylor and 
Francis Galton in England, men for whom he retained a profound 
respect, which more suo did not preclude critical dissent. Here 
again it would be instructive to learn more about the measure of 
their direct influence. Tylor's famous paper on the application of 
statistics to sociological problems (1889) certainly impressed 
Boas; for a while, he told me, it seemed as though everything 
could be solved by the methods there outlined. Galton he re- 
garded as the true father of biometrics, for which Karl Pearson 
had furnished the teclinical apparatus. He recognized, of course, 
Pearson's exceptional ability and once tried to visit him in Eng- 
land; but Pearson, though he had referred very cordially to Boas 
in the second edition of The Grammar of Science, for some reason 
declined to see him. 

In 1888 Boas accepted a docentship at Clark University, re- 
maining there until 1892, when he had his and America's first 
anthropological Ph.D. student, A. F. Chamberlain. He left Clark 

^BAE, 6th Ann. Rept. (1888). 


Theories and Theorists 

to become F. W. Putnam's chief assistant at the anthropological 
exhibits of the Chicago World's Fair, the core of the subsequent 
Field (Columbian) Museum. At this new institution he served as 
curator of anthropology, but was superseded by William H. 
Holmes. A year or two later he accepted an assistant curator- 
ship under Putnam at the American Museum of Natural History, 
a position soon combined with a lecturership at Columbia. At that 
time this institution offered anthropological work under several 
distinct auspices, Ripley of The Races of Europe fame lecturing 
on that subject in the department of economics, wliile Livingston 
Farrand held forth on comparative sociology, religion, and art 
in the department of psychology. In 1889, however, Boas was 
appointed to head a new department of antliropology, with Far- 
rand as his adjunct. Two years later he also became Putnam's 
successor at the American Museum. 

His dual responsibility enabled Boas to bring students into 
contact with anthropological collections and, above all, to provide 
them with opportunities for field work under the auspices of the 
Museum. During this period developed the most ambitious re- 
search project of his career, the Morris K. Jesup Expedition, ac- 
tually a series of expeditions designed to shed light on Asiatic- 
American relationships. Boas's collaborators included Farrand, 
Harlan I. Smith, and other Americans, as well as several noted 
European scholars, such as Waldemar Bogoras, Waldemar Jochel- 
son, and Berthold Laufer. In this connection and later Boas 
evinced a rare capacity for enlisting the cooperation of men 
qualified to advance science. It was during his curatorship, too, 
that Roland B. Dixon, assisted by A. M. Tozzer, undertook the 
first strictly scientific investigation of a Californian tribe, culmi- 
nating in the model monograph on The Northern Maidu. Even 
unacademic men — intelligent whalers, such as Captains Mutch 
and Comer — were drafted to make systematic observations on 
the Central Eskimo. 

Several students, subsequently distinguished in the science, won 
their ethnographer's spurs under Boas's jointly curatorial and pro- 
fessorial tutelage — A. L. Kroeber, Clark Wissler, Wilfiam Jones. 
Another fruitful institutional connection resulted from Boas's ap- 
pointment ( 1901 ) as Honorary Philologist of the Bureau of Ameri- 


Franz Boas, 1858-1942 

can Ethnology. It facilitated the accumulation and ultimate pub- 
lication of vast bodies of linguistic material, as evidenced in the 
Handbook of American Indian Languages (1911, 1922). 

A clash with Dr. Hermon C. Bumpus, then director of the Amer- 
ican Museum, concerning methods of installation and the generic 
issue of departmental autonomy, led to Boas's resignation ( 1905 ) 
as curator and for many years severed the intimate bonds of the 
Museum department with that at Columbia. However, he soon 
found other outlets for his surplus energy. In 1908 he became edi- 
tor of the Journal of American Folk-Lore; in 1910 he helped create 
the International School of American Archaeology and Ethnol- 
ogy in Mexico; in 1917 he founded the International Journal of 
American Linguistics; and for many years he edited the Publica- 
tions of the American Ethnological Society. In 1908, moreover, 
the United States Immigration Commission authorized him to 
undertake a somatological study of European immigrants. The 
task once more involved the careful planning of a large-scale 
project with the aid of many assistants. Nor did personal field 
work cease: he directed excavations in Mexico and Porto Rico, 
went to the Kootenay and to the Keresan Indians, even revisited 
the Kwakiutl in his old age. Besides all this he regularly attended 
scientific congresses in America and Europe. 

Boas's many-sided scientific activities found national and in- 
ternational recognition. He was elected to the National Academy 
of Sciences in April, 1900; was a member of the American Philo- 
sophical Society; president of the American Association for the 
Advancement of Science in 1931 and of the New York Academy 
of Sciences in 1910. Among his foreign honors may be mentioned 
the doctorate bestowed by Oxford University. 

World War I and its aftermath brought to the fore some little- 
suspected facets of Boas's personality. He had long acquired Amer- 
ican citizenship, but like many others found himself beset by a 
conflict of emotions. He was an internationalist if ever there was 
one; but he was also steeped in the culture of his native land, had 
close relatives living there, was linked by personal and profes- 
sional ties with innumerable Germans. What is more, he had been 
in his 'teens when the millennial dream of a united Reich had 
come true; had Uved through a period of spectacular positive 


Theories and Theorists 

achievement in Germany. His attitude could not well be that of 
the 'forty-eight immigrants. He himself was aware of the difference 
and — probably thinking of his uncle. Dr. A. Jacobi — alluded in 
conversation to this disagreement between the two generations 
of German-Americans. 

Feeling strongly, as always, on matters of principle, he bitterly 
resented the pro-Ally attitude of Americans as a breach of neu- 
trality. "Oh, if we had a Grover Cleveland in the White House!" 
he exclaimed once, presumably referring to the Venezuelan epi- 
sode. His was not a temperament that could restrain utterance in 
such a crisis. He wrote letters to the New York Times and to The 
Nation; contributed articles to The Dial and the Illinois Staats- 
zeitung; preached a sermon on internationalism at St. Clark's 
Church; and parried a move to investigate the loyalty of the Co- 
lumbia faculty by reading to a class his ideas on patriotism. 

To be sure, only misunderstanding or malice could construe 
his stand as nationalistically pro-Kaiser. His posthumously issued 
pronouncements dating back to the period '^ read very well in 
1946. But tliirty years ago they jarred upon people on the verge 
of war or actually embattled. Such auditors did not like to hear 
that their individualism and democracy, rooted in local condi- 
tions, were not necessarily superior to a polity like Germany's 
with her very different history. Still less did they thrill to the 
idea that obligations to mankind ought to take precedence of 
patriotism; nor were their susceptibilities assuaged by the admis- 
sion that patriots — like the witch-hunters of an earlier period — 
might be utterly sincere and morally pure. Boas, however, pur- 
sued his way, unmindful of general unpopularity and the threat 
of academic discipline. 

Post-bellum days saw him at the unprovocative but equally 
novel task of organizing the Emergency Society for German and 
Austrian Science. The man who had hitherto begrudged every 
minute of social life as an encroachment on professional work now 
lavished precious hours on routine jobs, on correspondence, on 
the search for new contacts that might aid in restoring the im- 
periled life of German science. This labor of love and self-abnega- 
tion was duly appreciated by its beneficiaries. When he applied 

^ Race and Democratic Society (New York, 1945). 


Franz Boas, 1858-1942 

for a visa in 1924, the German Consul at New York declined the 
customary fee. At the International Americanists' Congress held 
in Hamburg in 1930 Professor Sapper conveyed to Boas a diploma 
of honorary member sliip in the Geographical Society of Wiirz- 
burg, at the same time lauding his efforts on behalf of German 
scholarship, "for many a scientific post was able to resist the 
financial pressure of those days solely thanks to his organization 
of German-American aid." Similarly, Brockhaus' encyclopaedia 
celebrated his "grossziigige Organisation der Unterstiitzung der 
deutschen Wissenschaft." ^ 

Maintaining singular mental alertness, Boas remained at Co- 
lumbia long beyond the usual length of service. He became "emeri- 
tus in residence" in 1936, emeritus in 1938, but retained his old 
office at the University. As a septuagenarian he continued to loom 
large at international congresses, still made trips to Europe, and 
continued to inspire students and visitors from foreign parts. 
His declining years were fraught with sadness. He lost his younger 
son and his wife in automobile accidents, and his second daughter 
through an insidious disease. Such solace as was possible he found 
in unremitting scientific work. 

The rise of Hitler stirred him to the depths of his soul. That 
the country whose cultural heritage he gloried in, the country 
on whose behalf he had sufiFered abuse and ostracism in the 
first World War, should flout the principles dear to him was an 
unbearable thought. Besides, being of Jewish extraction, he had 
relatives in Germany whose very existence was threatened by the 
Umbruch. He reacted once more in character, writing an open 
letter to President Hindenburg, denouncing the tenets of Nazism 
in the daily press or in popular magazines; dragging himself when 
already enfeebled by old age and an encroaching heart disease 
to the platform at public gatherings in order to inveigh against 
Hitlerian excesses. 

His campaign against racism naturally brought him a wider 
following than the monographs on the Kwakiutl or even the aca- 
demic treatment of race in The Mind of Pritnitive Man (1911, 
1938). He became the spokesman not only of disinterested hu- 
manitarians, but also of Leftists and Communists. Communists 

^Vol. Ill (1929), 66. 


Theories and Theorists 

are not universally popular, and even in quarters averse to Nazi 
philosophy the association with them sufficed to make Boas a sus- 
pect fellow- traveler. The facts seem to be as follows. Boas had 
a live social sense that automatically made him favor the under- 
dog, so that he was unquestionably a liberal rather than a con- 
servative in his general outlook. On the other hand, he loathed 
regimentation, whether by a college president, a party machine, 
or an unenlightened public opinion. When Lily Braun, the rene- 
gade daughter of a Prussian general, published her memoirs in 
1908, Boas read them and was repelled by their picture of Social 
Democratic party tyranny. In possibly my very last conversation 
with him, a year or so before his death, he broached the subject 
of the Bolsheviks, summarizing his position in these words: "The 
Communists have done many very good things, they have also 
done many very bad things." Assuredly this was not the voice of 
blind partisanship. As for Marxist doctrines, he had all his life 
recoiled from closed systems, hence could not accept a philosophy 
of economic determinism or any other dogmatic scheme. On the 
other hand, no one was less likely than he to avoid contacts simply 
because they might arouse general disapproval. "Fellow-traveler," 
"pink," and "red" were to him meaningless catchwords. 

On December 21, 1942, Boas was lunching with Professor 
Paul Rivet (Paris) at the Columbia Faculty Club in the com- 
pany of several colleagues. The guest of honor has graphically 
recorded the experience.*^ Boas had just voiced his contempt for 
racism, when the fatal stroke occurred: "Sans un cri, sans une 
plainte, nous le vimes se renverser en arriere; quelques rales, un 
grand cerveau avait cesse de penser." 

Boas's services to anthropology were so great and manifold that 
occasionally enthusiastic disciples unfamiliar with history talked 
and wrote as though his predecessors and contemporaries were 
negligible. One obituary article declared: "He found anthropology 
a collection of wild guesses and a happy hunting ground for the 
romantic lover of primitive things; he left it a discipline in which 
theories could be tested and in which he had delimited possibilities 
from impossibilities." This is to parade Boas as a mytliological 

*Paul Rivet, "Franz Boas," Renaissance, I (1943), 313 f. 


Franz Boas, 1858-1942 

culture-hero creating something out of nothing. The conception 
would have been intolerable to Boas, who fully esteemed what 
had been done by E. B. Tylor, Lewis H. Morgan, Eduard Hahn, 
Karl von den Steinen, and others. Indeed, he was especially ap- 
preciative of men who had achieved what he himself never at- 
tempted — an intimate, yet authentic, picture of aboriginal life. I 
have hardly ever heard him speak with such veritable enthusiasm 
as when lauding Bogoras' account of the Chukchi, Rasmussen's 
of the Eskimo, Turi's of the Lapps. 

In the following paragraphs, then, I shall try to sketch Boas's 
achievement in perspective and without unfairness to others. 

To begin with an obvious fact, he approached the study of man 
from every angle: as Rivet puts it, "son oeuvre embrasse le prob- 
leme humain dans son entier . . . Tout ce qui concerne I'homme 
sollicite sa curiosite . . ." What is more, from the start he saw the 
need of acquiring in each branch of the science the highest degree 
of technical equipment. The physical anthropologist must use the 
tools of biometrics; the linguist must become a phonetician and 
an analyst along the lines of Indo-European philology; the eth- 
nographer must envisage the subtler as well as the more obvious 
phases of social life — folk-literature, music, the subjective atti- 
tudes of primitive man no less than artifacts or social structure. 
Nothing is more remarkable than the systematic way in which 
Boas, trained in quite different fields, acquired the techniques 
requisite for the highest type of work in the several subdivisions 
of anthropology. Even in archaeology, which he treated with 
comparative neglect, his work has been declared to show "a per- 
fect appreciation of the problems and the best archaeological 

Further, this many-sided virtuosity was justified by the solidity 
of his results. Everywhere he saw new problems and devised new 
methods of attack. Even his archaeological contributions. Mason 
assures us, "all have been substantiated by later and more de- 
tailed work. They have formed the basis for all later research in 
this region." What is more, they preceded by several years the 
stratigraphic approach that rightly shed luster on Kidder's and 
Nelson's work in the Pueblo area. Again, in linguistics, Boas was, 


Theories and Theorists 

if not the first, yet the most persistent "to analyze exotic material 
without forcing it into the strait-jacket of the familiar" ( Emeneau ) . 
As a physical anthropologist he deprecated sheer taxonomy; de- 
fined race on a profounder basis; demonstrated the {nota bene, 
limited) plasticity of the human organism; studied the phenomena 
of growth on a major scale; and was one of the earliest investiga- 
tors to note segregation in hybrid human groups. His ethnological 
contributions were so varied that two must suffice for purposes 
of illustration. He was the first to inquire into the aboriginal art- 
ist's subjective attitude toward his tasks; and, paralleling the work 
of Homeric scholars, he correlated the social life depicted in a 
people's folk-literature with their observed culture. In theory 
he may be described as an epistemologist rather than a meta- 
physician: he suspected traditional labels and catchwords, in- 
quired into their empirical foundation, and often arrived at a new 
and illuminating re-classification of data. 

Tastes differ in science, as in everything else. Hence Boas's 
achievement was bound to disappoint certain minds. Keenly 
aware of the gaps in our knowledge, he refused to fill them with 
plausible speculations resulting in a spuriously complete picture 
of the whole field. He proclaimed no all-bracing "laws" and, ex- 
cept for his views on race, voiced no simple message that might 
appeal to large masses. In point of form he lacked the polished 
diction of a Frazer or the sprightly humor of his friend, Karl von 
den Steinen. Nor did he complete a single large-scale portrait of 
a tribal culture, not even of his beloved Kwakiutl. 

Similarly, his teaching was not designed for everyone's palate. 
The most effective trainer of anthropological investigators was 
not an ideal pedagogue. He was, indeed, uncanny in his capacity 
to harness a student's skills for the advancement of science, but 
he did not trouble to ferret out a learner's needs at a particular 
stage of progress. Novices were not pampered with milk for babes. 
Fearful lest they turn dilettanti, he imposed on virtually every new- 
comer in my day his course on statistical theory (usually audited 
by professors from other departments ) and another on American 
Indian languages. His ethnographic lectures rarely, if ever, sys- 
tematically surveyed the area announced, but discussed the prob- 
lems that engaged his attention. Other men's views he often 


Franz Boas, 1858-1942 

treated in a way likely to mislead the immature, for by concen- 
trating on controversial issues he sometimes conveyed the im- 
pression of total condemnation when there was merely partial 
dissent. One might easily carry away the idea that he had a low 
opinion of Tylor or Ratzel, as was certainly not the case. His cri- 
tique of environmentalism, for instance, was urged so forcibly 
that for years I failed to grasp how carefully he took cognizance 
of geographical factors. As to the skepticism he instilled by pre- 
cept and example, he himself was at times smitten with qualms, 
wondering whether he was inhibiting the free play of the imagina- 
tion, which, contrary to appearances, he rated very high. One 
student summarized his total reaction after a seminar of Boas's as 
follows: "All books are bad, articles may be good," the suppressed 
implication being that even they seldom were. 

Yet he valued high-class work even when done by men of 
utterly diflFerent personality. Of Bogoras and Rasmussen I have 
already spoken. He keenly appreciated Francis Galton, William 
James, William Morton Wheeler, Karl von den Steinen. Of his 
Columbia colleagues I think he rated E. B. Wilson highest. 
"He is a first-rate man," he once said to me. Thomas Hunt Morgan 
he accepted as "very good," but with qualifications. Among Wash- 
ington scientists, Karl Grove Gilbert enjoyed his esteem. Contrary 
to opinions occasionally heard, his scientific judgment was little 
warped by personal animosity. There was not much love lost be- 
tween him and certain Washingtonian colleagues; but he de- 
scribed one of his bitterest enemies to me as a man of great native 
ability and gave another full credit for founding a technical 

To revert to his teaching, my novitiate probably came at the 
worst possible period for establishing rapport, for it was the time 
of his feud with the director of the American Museum. Boas 
seemed perpetually busy and preoccupied. I actually dreaded 
meeting him on the way to classes in Schermerhom Hall. Utter 
silence would follow a curt "Good morning" till I found the situa- 
tion intolerable. "Have you read Kollmann's article on Pygmies in 
the last issue of Globus?" I once asked him on one of these em- 
barrassing occasions. He answered, "No"; I offered a few remarks 
on the subject; then we again walked on in silence. Having a con- 


Theories and Theorists 

ference with him was something of an event for A. B. Lewis, A. A. 
Goldenweiser, Paul Radin, and myself; Speck and Sapir, with 
their philological background, enjoyed, I think, a rather easier 
entree. This also had held true in Kroeber's and William Jones's 
time, and as a visitor in later periods I was able to watch his free 
and easy relations with subsequent generations of his disciples. 
Accordingly, I cannot but ascribe his earlier reserve to the tribu- 
lations of the era. 

Systematic information, as indicated, he did not vouchsafe in 
ethnological courses, that the student was supposed somehow to 
get for himself. Yet it was not an easy task at a time when the 
good books had grown antiquated, so that trustworthy knowledge 
was obtainable only by wading through tomes of unilluminating 
descriptive detail. However, Boas was singularly unexacting in 
regard to a student's factual information. Probably there is not 
nowadays a single undergraduate major in any of our large an- 
thropological departments who does not control a wider range 
of data than I did when Boas deemed me fit for the doctorate. 
It was enough that I had worked in the field, gained a theoretical 
conception there, and tlirashed out the issue in a formal paper. 
On the other hand, he came very near holding up A. B. Lewis, 
whose knowledge was incomparably superior to mine, but whose 
dissertation discussed nothing of theoretical significance. Berthold 
Laufer, who liked it, observed querulously to me, "Boas always 
wants a thesis to have a point!" 

Why did we reverence so indifferent a pedagogue as a great 
teacher? For the same reason, no doubt, that in later years mature 
men and women — Elsie Clews Parsons, Pliny Earle Goddard, and 
George A. Dorsey, for example — hailed him as their leader. Yet 
Goddard had come to New York full of skepticism about Boas; 
and Dorsey had been at swords' points with him in the American 
Anthropological Association. The explanation is simple. Here was 
a scientist primarily interested in science — not in the organization 
of research, not in the personalities of colleagues, not in a display 
of his cleverness, but in the problems that sprang from his data, 
in the quest of the truth. He seemed to personify the very spirit 
of science, and with his high seriousness — unsurpassed by any in- 
vestigator I have known in any sphere — he communicated some- 


Franz Boas, 1858-1942 

thing of that spirit to others. Therein Hes his greatness as a teacher. 

Constituted as he was, he could not avoid misunderstandings 
either as to his views or his character. Even scientific guilds live 
by slogans and balk at finer distinctions. Boas threw out a hint 
how totemism might have evolved in British Columbia and was 
forthwith credited with a universal theory of the phenomenon. 
Pointing to the positive achievements of colored races, he re- 
jected the arguments of racists, hence was either hailed or de- 
nounced as a dogmatic equalitarian. Yet he clearly formulated 
in both editions of his most popular book a rather different posi- 
tion: "It may be well to state here once more with some emphasis 
that it would be erroneous to claim as proved that there are no 
differences in the mental make-up of the Negro race taken as a 
whole and of any other race taken as a whole, and that their ac- 
tivities should run in exactly the same lines." ^ Again, his cham- 
pionship of a strictly limited plasticity was misinterpreted as a 
denial of heredity. Some forty years ago, at a joint meeting of 
anthropologists and psychologists, even h